Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
We train open frontier models and ship the same stack to our customers. Its spans the full stack of training, deploying and continuously improving models — compute, large-scale RL, environments, sandboxes, evals, and deployment.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Semianalysis, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Build and optimize the distributed training infrastructure behind our pre-training and large-scale RL training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.
Build and optimize the distributed training infrastructure behind our pre-training and large-scale RL training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.
Improve end-to-end training efficiency across compute, memory, networking, and scheduling layers.
Design and implement low-level performance optimizations, including kernels, communication paths, and runtime improvements.
Work on distributed training systems spanning data, tensor, and pipeline parallel workloads.
Help shape the architecture of our RL training stack, including async rollout and post-training systems.
Contribute to open-source libraries and internal infrastructure used for frontier-scale model training.
Collaborate closely with researchers and infrastructure engineers to translate bottlenecks into concrete systems improvements.
Stay at the frontier of training systems, inference systems, compiler/runtime tooling, and hardware-aware optimization techniques.
Strong systems engineering experience in AI/ML infrastructure, especially around large-scale model training or inference.
Deep familiarity with PyTorch and distributed training frameworks such as PyTorch Distributed, DeepSpeed, FSDP, Megatron, vLLM, Ray, or related tooling.
Experience optimizing training performance across kernels, memory movement, communication overhead, or parallelization strategy.
Hands-on experience with large-scale training techniques including data parallelism, tensor parallelism, and pipeline parallelism.
Strong understanding of GPU architecture, profiling, and performance debugging.
Ability to identify bottlenecks across the stack and drive improvements from first principles.
Comfort working in a fast-moving environment with ambiguous problems and high ownership.
Especially Exciting (optional)
Experience writing or optimizing CUDA / Triton kernels.
Experience with compiler or runtime optimization for ML systems.
Experience working on RL training infrastructure, rollout systems, or asynchronous training pipelines.
Experience with multi-node GPU clusters and high-performance networking.
Contributions to open-source ML systems or infrastructure projects.
Interest in publishing technical work or sharing insights through engineering blogs and technical writing.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you’re excited about building the systems foundation for frontier-scale training and open superintelligence, we’d love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nWe train open frontier models and ship the same stack to our customers. Its spans the full stack of training, deploying and continuously improving models — compute, large-scale RL, environments, sandboxes, evals, and deployment.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Semianalysis, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n \n\n\nROLE IMPACT\n\nBuild and optimize the distributed training infrastructure behind our pre-training and large-scale RL training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.\n\n\nRESPONSIBILITIES\n\n - Build and optimize the distributed training infrastructure behind our pre-training and large-scale RL training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.\n\n - Improve end-to-end training efficiency across compute, memory, networking, and scheduling layers.\n\n - Design and implement low-level performance optimizations, including kernels, communication paths, and runtime improvements.\n\n - Work on distributed training systems spanning data, tensor, and pipeline parallel workloads.\n\n - Help shape the architecture of our RL training stack, including async rollout and post-training systems.\n\n - Contribute to open-source libraries and internal infrastructure used for frontier-scale model training.\n\n - Collaborate closely with researchers and infrastructure engineers to translate bottlenecks into concrete systems improvements.\n\n - Stay at the frontier of training systems, inference systems, compiler/runtime tooling, and hardware-aware optimization techniques.\n\n\nREQUIREMENTS\n\n - Strong systems engineering experience in AI/ML infrastructure, especially around large-scale model training or inference.\n\n - Deep familiarity with PyTorch and distributed training frameworks such as PyTorch Distributed, DeepSpeed, FSDP, Megatron, vLLM, Ray, or related tooling.\n\n - Experience optimizing training performance across kernels, memory movement, communication overhead, or parallelization strategy.\n\n - Hands-on experience with large-scale training techniques including data parallelism, tensor parallelism, and pipeline parallelism.\n\n - Strong understanding of GPU architecture, profiling, and performance debugging.\n\n - Ability to identify bottlenecks across the stack and drive improvements from first principles.\n\n - Comfort working in a fast-moving environment with ambiguous problems and high ownership.\n\nEspecially Exciting (optional)\n\n - Experience writing or optimizing CUDA / Triton kernels.\n\n - Experience with compiler or runtime optimization for ML systems.\n\n - Experience working on RL training infrastructure, rollout systems, or asynchronous training pipelines.\n\n - Experience with multi-node GPU clusters and high-performance networking.\n\n - Contributions to open-source ML systems or infrastructure projects.\n\n - Interest in publishing technical work or sharing insights through engineering blogs and technical writing.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you’re excited about building the systems foundation for frontier-scale training and open superintelligence, we’d love to hear from you."},{"id":"b9343f5d-326d-4b01-b6a7-acf56cffe3ac","title":"Open Application for Unconventional Talent","department":"Others","team":"Others","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-07-08T18:43:13.470+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/b9343f5d-326d-4b01-b6a7-acf56cffe3ac","applyUrl":"https://jobs.ashbyhq.com/primeintellect/b9343f5d-326d-4b01-b6a7-acf56cffe3ac/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
If you’ve built complex technical systems, contributed meaningfully to open-source projects, or mastered multiple domains, we want to hear from you. Whether your expertise is in AI, distributed computing, cryptography, systems programming, or something unexpected, what matters is your ability to learn fast, think rigorously, and execute.
Tell us what excites you about PrimeIntellect, something impressive that that you’ve built, and how you’d accelerate open and decentralized AGI.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team. \n\n\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\n\nIf you’ve built complex technical systems, contributed meaningfully to open-source projects, or mastered multiple domains, we want to hear from you. Whether your expertise is in AI, distributed computing, cryptography, systems programming, or something unexpected, what matters is your ability to learn fast, think rigorously, and execute.\n\nTell us what excites you about PrimeIntellect, something impressive that that you’ve built, and how you’d accelerate open and decentralized AGI."},{"id":"abfa70f7-a6f1-44d2-a6c1-560e1c8477d4","title":"Member of Technical Staff - Inference","department":"Engineering","team":"Engineering","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-07-08T18:45:28.420+00:00","isListed":true,"isRemote":true,"workplaceType":"Remote","address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/abfa70f7-a6f1-44d2-a6c1-560e1c8477d4","applyUrl":"https://jobs.ashbyhq.com/primeintellect/abfa70f7-a6f1-44d2-a6c1-560e1c8477d4/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
This is a hybrid position spanning cloud LLM serving, LLM inference optimization and RL systems. You will be working on advancing our ability to evaluate and serve models trained with our RL Lab at scale. The two key areas are:
Building the infrastructure to serve LLMs efficiently at scale.
Optimization and integration of inference systems into our RL training stack.
LLM Serving
Multi‑tenant LLM Serving: Build a multi-tenant LLM serving platform that operates across our cloud GPU fleets.
GPU‑Aware Scheduling: Design placement and scheduling algorithms for heterogeneous accelerators.
Resilience & Failover: Implement multi‑region/zone failover and traffic shifting for resilience and cost control.
Autoscaling & Routing: Build autoscaling, routing, and load balancing to meet throughput/latency SLOs.
Model Distribution: Optimize model distribution and cold-start times across clusters.
Inference Optimization & Performance
Framework Development: Integrate and contribute to LLM inference frameworks such as vLLM, SGLang, TensorRT‑LLM.
Parallelism and Configuration Tuning: Optimize configurations for tensor/pipeline/expert parallelism, prefix caching, memory management and other axes for maximum performance.
End‑to‑End Performance: Profile kernels, memory bandwidth and transport; apply techniques such as quantization and speculative decoding.
Perf Suites: Develop reproducible performance suites (latency, throughput, context length, batch size, precision).
RL Integration: Embed and optimize distributed inference within our RL stack.
Platform & Tooling
CI/CD: Establish CI/CD with artifact promotion, performance gates, and reproducible builds.
Observability: Build metrics, logs, tracing; structured incident response and SLO management.
Docs & Collaboration: Document architectures, playbooks, and API contracts; mentor and collaborate cross‑functionally.
Required Experience
Building ML Systems at Scale: 3+ years building and running large‑scale ML/LLM services with clear latency/availability SLOs.
Inference Backends: Hands‑on with at least one of vLLM, SGLang, TensorRT‑LLM.
Distributed Serving Infra: Familiarity with distributed and disaggregated serving infrastructure such as NVIDIA Dynamo.
Inference Internals: Deep understanding of prefill vs. decode, KV‑cache behavior, batching, sampling, speculative decoding, parallelism strategies.
Full‑Stack Debugging: Comfortable debugging CUDA/NCCL, drivers/kernels, containers, service mesh/networking, and storage, owning incidents end‑to‑end.
Infrastructure Skills
Python: Systems tooling and backend services.
PyTorch: LLM Inference engine development and integration, deployment readiness.
Cloud & Automation: AWS/GCP service experience, cloud deployment patterns.
Kubernetes: Running infrastructure at scale with containers on Kubernetes.
GPU & Networking: Architecture, CUDA runtime, NCCL, InfiniBand; GPU‑aware bin‑packing and scheduling across heterogeneous fleets.
Nice to Have
Kernel‑Level Optimization: Familiarity with CUDA/Triton kernel development; Nsight Systems/Compute profiling.
Systems Performance Languages: Rust, C++.
Data & Observability: Kafka/PubSub, Redis, gRPC/Protobuf; Prometheus/Grafana, OpenTelemetry; reliability patterns.
Infra & Config Automation: Terraform/Ansible, infrastructure-as-code, reproducible environments
Open Source: Contributions to serving, inference, or RL infrastructure projects.
Cash Compensation Range of $150-300k with significant equity incentives
Flexible work arrangement (remote or San Francisco office)
Full visa sponsorship and relocation support
Professional development budget
Regular team off-sites and conference attendance
Opportunity to shape decentralized AI and RL at Prime Intellect
You'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source contributions.
We value potential over perfection. If you're passionate about democratizing AI development, we want to talk to you.
Ready to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team. \n\n\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\n\n\nROLE IMPACT\n\nThis is a hybrid position spanning cloud LLM serving, LLM inference optimization and RL systems. You will be working on advancing our ability to evaluate and serve models trained with our RL Lab at scale. The two key areas are:\n\n 1. Building the infrastructure to serve LLMs efficiently at scale.\n\n 2. Optimization and integration of inference systems into our RL training stack.\n\n\n\n\nCORE TECHNICAL RESPONSIBILITIES\n\nLLM Serving\n\n - Multi‑tenant LLM Serving: Build a multi-tenant LLM serving platform that operates across our cloud GPU fleets.\n\n - GPU‑Aware Scheduling: Design placement and scheduling algorithms for heterogeneous accelerators.\n\n - Resilience & Failover: Implement multi‑region/zone failover and traffic shifting for resilience and cost control.\n\n - Autoscaling & Routing: Build autoscaling, routing, and load balancing to meet throughput/latency SLOs.\n\n - Model Distribution: Optimize model distribution and cold-start times across clusters.\n\nInference Optimization & Performance\n\n - Framework Development: Integrate and contribute to LLM inference frameworks such as vLLM, SGLang, TensorRT‑LLM.\n\n - Parallelism and Configuration Tuning: Optimize configurations for tensor/pipeline/expert parallelism, prefix caching, memory management and other axes for maximum performance.\n\n - End‑to‑End Performance: Profile kernels, memory bandwidth and transport; apply techniques such as quantization and speculative decoding.\n\n - Perf Suites: Develop reproducible performance suites (latency, throughput, context length, batch size, precision).\n\n - RL Integration: Embed and optimize distributed inference within our RL stack.\n\nPlatform & Tooling\n\n - CI/CD: Establish CI/CD with artifact promotion, performance gates, and reproducible builds.\n\n - Observability: Build metrics, logs, tracing; structured incident response and SLO management.\n\n - Docs & Collaboration: Document architectures, playbooks, and API contracts; mentor and collaborate cross‑functionally.\n\n\nTECHNICAL REQUIREMENTS\n\nRequired Experience\n\n - Building ML Systems at Scale: 3+ years building and running large‑scale ML/LLM services with clear latency/availability SLOs.\n\n - Inference Backends: Hands‑on with at least one of vLLM, SGLang, TensorRT‑LLM.\n\n - Distributed Serving Infra: Familiarity with distributed and disaggregated serving infrastructure such as NVIDIA Dynamo.\n\n - Inference Internals: Deep understanding of prefill vs. decode, KV‑cache behavior, batching, sampling, speculative decoding, parallelism strategies.\n\n - Full‑Stack Debugging: Comfortable debugging CUDA/NCCL, drivers/kernels, containers, service mesh/networking, and storage, owning incidents end‑to‑end.\n\nInfrastructure Skills\n\n - Python: Systems tooling and backend services.\n\n - PyTorch: LLM Inference engine development and integration, deployment readiness.\n\n - Cloud & Automation: AWS/GCP service experience, cloud deployment patterns.\n\n - Kubernetes: Running infrastructure at scale with containers on Kubernetes.\n\n - GPU & Networking: Architecture, CUDA runtime, NCCL, InfiniBand; GPU‑aware bin‑packing and scheduling across heterogeneous fleets.\n\nNice to Have\n\n - Kernel‑Level Optimization: Familiarity with CUDA/Triton kernel development; Nsight Systems/Compute profiling.\n\n - Systems Performance Languages: Rust, C++.\n\n - Data & Observability: Kafka/PubSub, Redis, gRPC/Protobuf; Prometheus/Grafana, OpenTelemetry; reliability patterns.\n\n - Infra & Config Automation: Terraform/Ansible, infrastructure-as-code, reproducible environments\n\n - Open Source: Contributions to serving, inference, or RL infrastructure projects.\n\n\nWHAT WE OFFER\n\n - Cash Compensation Range of $150-300k with significant equity incentives\n\n - Flexible work arrangement (remote or San Francisco office)\n\n - Full visa sponsorship and relocation support\n\n - Professional development budget\n\n - Regular team off-sites and conference attendance\n\n - Opportunity to shape decentralized AI and RL at Prime Intellect\n\n\nGROWTH OPPORTUNITY\n\nYou'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source contributions.\n\nWe value potential over perfection. If you're passionate about democratizing AI development, we want to talk to you.\n\nReady to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone."},{"id":"297d925e-5a42-40bd-b02f-5c928d226f18","title":"Member of Technical Staff - GPU Infrastructure","department":"Engineering","team":"Engineering","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"APAC","address":null},{"location":"United States","address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}}}],"publishedAt":"2026-07-08T18:45:18.934+00:00","isListed":true,"isRemote":true,"workplaceType":"Remote","address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/297d925e-5a42-40bd-b02f-5c928d226f18","applyUrl":"https://jobs.ashbyhq.com/primeintellect/297d925e-5a42-40bd-b02f-5c928d226f18/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
We're hiring for multiple GPU Infrastructure positions: one based in the APAC region and one based in the USA.
This customer-facing role combines deep technical expertise with hands-on implementation. You'll be instrumental in:
Partner with clients to understand workload requirements and design optimal GPU cluster architectures
Create technical proposals and capacity planning for clusters ranging from 100 to 10,000+ GPUs
Develop deployment strategies for LLM training, inference, and HPC workloads
Present architectural recommendations to technical and executive stakeholders
Deploy and configure orchestration systems including SLURM and Kubernetes for distributed workloads
Implement high-performance networking with InfiniBand, RoCE, and NVLink interconnects
Optimize GPU utilization, memory management, and inter-node communication
Configure parallel filesystems (Lustre, BeeGFS, GPFS) for optimal I/O performance
Tune system performance from kernel parameters to CUDA configurations
Serve as primary technical escalation point for customer infrastructure issues
Diagnose and resolve complex problems across the full stack - hardware, drivers, networking, and software
Implement monitoring, alerting, and automated remediation systems
Provide 24/7 on-call support for critical customer deployments
Create runbooks and documentation for customer operations teams
3+ years hands-on experience with GPU clusters and HPC environments
Deep expertise with SLURM and Kubernetes in production GPU settings
Proven experience with InfiniBand configuration and troubleshooting
Strong understanding of NVIDIA GPU architecture, CUDA ecosystem, and driver stack
Experience with infrastructure automation tools (Ansible, Terraform)
Proficiency in Python, Bash, and systems programming
Track record of customer-facing technical leadership
NVIDIA driver installation and troubleshooting (CUDA, Fabric Manager, DCGM)
Container runtime configuration for GPUs (Docker, Containerd, Enroot)
Linux kernel tuning and performance optimization
Network topology design for AI workloads
Power and cooling requirements for high-density GPU deployments
Experience with 1000+ GPU deployments
NVIDIA DGX, HGX, or SuperPOD certification
Distributed training frameworks (PyTorch FSDP, DeepSpeed, Megatron-LM)
ML framework optimization and profiling
Experience with AMD MI300 or Intel Gaudi accelerators
Contributions to open-source HPC/AI infrastructure projects
Cash compensation of $150-300k with significant equity incentives
Flexible work arrangement (remote or San Francisco office)
Full visa sponsorship and relocation support
A professional development budget for courses and conferences
Regular team off-sites and conference attendance
The opportunity to shape core infrastructure for frontier AI research
You'll work directly with customers pushing the boundaries of AI, from startups training foundation models to enterprises deploying massive inference infrastructure. You'll collaborate with our world-class engineering team while having direct impact on systems powering the next generation of AI breakthroughs.
We value expertise and customer obsession - if you're passionate about building reliable, high-performance GPU infrastructure and have a track record of successful large-scale deployments, we want to talk to you.
Apply now and join us in our mission to democratize access to planetary scale computing.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\nWe're hiring for multiple GPU Infrastructure positions: one based in the APAC region and one based in the USA.\n\n\nCORE TECHNICAL RESPONSIBILITIES\n\nThis customer-facing role combines deep technical expertise with hands-on implementation. You'll be instrumental in:\n\n\nCUSTOMER ARCHITECTURE & DESIGN\n\n - Partner with clients to understand workload requirements and design optimal GPU cluster architectures\n\n - Create technical proposals and capacity planning for clusters ranging from 100 to 10,000+ GPUs\n\n - Develop deployment strategies for LLM training, inference, and HPC workloads\n\n - Present architectural recommendations to technical and executive stakeholders\n\n\nINFRASTRUCTURE DEPLOYMENT & OPTIMIZATION\n\n - Deploy and configure orchestration systems including SLURM and Kubernetes for distributed workloads\n\n - Implement high-performance networking with InfiniBand, RoCE, and NVLink interconnects\n\n - Optimize GPU utilization, memory management, and inter-node communication\n\n - Configure parallel filesystems (Lustre, BeeGFS, GPFS) for optimal I/O performance\n\n - Tune system performance from kernel parameters to CUDA configurations\n\n\nPRODUCTION OPERATIONS & SUPPORT\n\n - Serve as primary technical escalation point for customer infrastructure issues\n\n - Diagnose and resolve complex problems across the full stack - hardware, drivers, networking, and software\n\n - Implement monitoring, alerting, and automated remediation systems\n\n - Provide 24/7 on-call support for critical customer deployments\n\n - Create runbooks and documentation for customer operations teams\n\n\nTECHNICAL REQUIREMENTS\n\n\nREQUIRED EXPERIENCE\n\n - 3+ years hands-on experience with GPU clusters and HPC environments\n\n - Deep expertise with SLURM and Kubernetes in production GPU settings\n\n - Proven experience with InfiniBand configuration and troubleshooting\n\n - Strong understanding of NVIDIA GPU architecture, CUDA ecosystem, and driver stack\n\n - Experience with infrastructure automation tools (Ansible, Terraform)\n\n - Proficiency in Python, Bash, and systems programming\n\n - Track record of customer-facing technical leadership\n\n\nINFRASTRUCTURE SKILLS\n\n - NVIDIA driver installation and troubleshooting (CUDA, Fabric Manager, DCGM)\n\n - Container runtime configuration for GPUs (Docker, Containerd, Enroot)\n\n - Linux kernel tuning and performance optimization\n\n - Network topology design for AI workloads\n\n - Power and cooling requirements for high-density GPU deployments\n\n\nNICE TO HAVE\n\n - Experience with 1000+ GPU deployments\n\n - NVIDIA DGX, HGX, or SuperPOD certification\n\n - Distributed training frameworks (PyTorch FSDP, DeepSpeed, Megatron-LM)\n\n - ML framework optimization and profiling\n\n - Experience with AMD MI300 or Intel Gaudi accelerators\n\n - Contributions to open-source HPC/AI infrastructure projects\n\n\nWHAT WE OFFER\n\n - Cash compensation of $150-300k with significant equity incentives\n\n - Flexible work arrangement (remote or San Francisco office)\n\n - Full visa sponsorship and relocation support\n\n - A professional development budget for courses and conferences\n\n - Regular team off-sites and conference attendance\n\n - The opportunity to shape core infrastructure for frontier AI research\n\n\nGROWTH OPPORTUNITY\n\nYou'll work directly with customers pushing the boundaries of AI, from startups training foundation models to enterprises deploying massive inference infrastructure. You'll collaborate with our world-class engineering team while having direct impact on systems powering the next generation of AI breakthroughs.\n\nWe value expertise and customer obsession - if you're passionate about building reliable, high-performance GPU infrastructure and have a track record of successful large-scale deployments, we want to talk to you.\n\nApply now and join us in our mission to democratize access to planetary scale computing."},{"id":"4c6ccfd6-6512-4905-979a-6dcc6d64a935","title":"Head of Growth","department":"Growth","team":"Growth","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-07-08T18:41:56.858+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/4c6ccfd6-6512-4905-979a-6dcc6d64a935","applyUrl":"https://jobs.ashbyhq.com/primeintellect/4c6ccfd6-6512-4905-979a-6dcc6d64a935/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
As Head of Growth, you’ll lead everything that connects our technology to the market — spanning sales, marketing, partnerships, and customer success. You’ll define how Prime Intellect tells its story, close large-scale post-training and compute contracts, and build the systems that let us scale revenue efficiently. This is a high-leverage role combining strategy, execution, and team building — perfect for a hybrid commercial operator who thrives at the intersection of infrastructure and intelligence.
Strategy & Vision
Build and lead a cross-functional Growth organization spanning Sales, Marketing, Partnerships, and Customer Success.
Define our go-to-market strategy for RL infrastructure pricing, packaging, and positioning.
Own revenue forecasting, pipeline visibility, and operational cadence (CRM, dashboards, and forecasting systems).
Design data-driven systems for deal tracking, reporting, and forecasting accuracy.
Sales & Partnerships
Lead enterprise sales cycles for post-training and multi-node cluster deals (64+ GPUs, 6+ month durations).
Develop repeatable playbooks for design-partner conversions and enterprise onboarding.
Build strategic partnerships with compute providers, AI companies, and research labs to expand our ecosystem.
Identify and nurture high-value leads through targeted outbound and inbound growth strategies.
Marketing
Drive developer and enterprise awareness through campaigns, launches, community, and content.
Translate technical features into clear market narratives and differentiated positioning.
Partner with Product to align market feedback with roadmap and developer onboarding.
Run quantitative experiments on activation, conversion, and usage metrics.
Customer Success / Account Management
Build the CS motion across onboarding, usage analytics, renewals, and expansion workflows.
Implement CRM-based renewal tracking and feedback loops into product and operations.
Establish systems for continuous measurement of customer value and satisfaction.
Growth Operations
Design scalable processes for deal management, supplier bidding, contract ops, and revenue tracking.
Partner with Operations and Finance on budgeting, hiring, and legal structures that support rapid scale.
Implement sales technology stack and CRM architecture to enable data-driven decision making.
5+ years leading growth, GTM, or revenue functions at a B2B or infrastructure startup.
Fluent in both technical and commercial conversations — comfortable discussing GPUs, APIs, and pricing models.
Proven ability to build scalable systems (CRM, forecasting, dashboards) and clean operational processes.
Track record of closing enterprise-scale infrastructure or developer-tooling deals.
Analytical, curious, and comfortable in high-ambiguity, fast-moving environments.
Strong communication and leadership skills with experience building small, high-performing teams.
Bonus:
Experience pricing or selling infrastructure or developer tools.
Familiarity with enterprise AI, cloud infrastructure, or post-training markets.
Early-stage operator mindset — bias for building playbooks, not inheriting them.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (remote or San Francisco)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
Opportunity to Shape Decentralized AI at Prime Intellect
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You'll help build our hosted training platform - the product that lets users launch LoRA and full fine-tuning runs on managed GPU clusters with a single API call or a few clicks. The role spans the developer-facing platform and the underlying Kubernetes-based training infrastructure that runs the jobs.
Design and operate Kubernetes-based training and inference orchestration across multi-cluster, multi-cloud GPU fleets
Build and maintain Helm charts that compose trainers, inference servers, environment servers, and supporting services into reproducible \"Training stacks\"
Develop the Python control-plane agents that watch pods, report run state to the platform, and keep clusters in sync
Implement scheduling and autoscaling for heterogeneous hardware (H100/H200/B200) using KEDA, LeaderWorkerSet, taints/tolerations, and gang scheduling
Run a tight GitOps workflow - every change ships through PRs, Helm values, and CI
Build node-local model caches, checkpoint pipelines, and shared storage for fast cold starts
Operate the observability stack (Prometheus, Grafana, Loki, DCGM) and make GPU cluster debugging fast
Build the developer-facing surfaces for hosted training: job submission, live run monitoring, logs, metrics, model/adapter management, comparisons
Develop FastAPI backend services and REST APIs that bridge the platform to running clusters
Build real-time monitoring and debugging tools (streaming logs, step-level metrics, failure analysis)
Ship product UI in Next.js / React / TypeScript with shadcn, Tailwind, tRPC, and TanStack Query
Interface with the RL trainer, inference servers, and environment servers running inside our clusters
Productize new training capabilities (new model architectures, RL algorithms, modes)
We're looking for engineers who are fluent across three areas - you don't need to be the world's best at any one, but you should have real depth in all three and a clear point of view on how they connect.
Strong working knowledge of the modern AI stack - open model families, finetuning techniques (LoRA, QLoRA, full FT, RLHF/RLAIF), inference engines (vLLM, SGLang, TensorRT-LLM)
Familiarity with GPU hardware tradeoffs (H100 / H200 / B200, NVLink, interconnects, memory hierarchy) and what they mean for training and inference workloads
Understanding of distributed training fundamentals (data/tensor/pipeline/expert parallelism, NCCL, multi-node scheduling)
Awareness of what's happening at the frontier - new models, training methods, infra patterns - and the ability to translate that into product decisions
Strong Kubernetes operations experience - Helm, CRDs, operators, KEDA, gang scheduling, GPU operator
Comfortable debugging real production clusters (kubectl, pod lifecycle, node issues, networking)
Cloud platform experience (GCP preferred - GCS, GKE, Cloud Run, Cloud Tasks)
Infrastructure automation (Helm, Terraform, Ansible) and a GitOps mindset
Observability: Prometheus, Grafana, Loki, OpenTelemetry, DCGM
Linux fundamentals: networking, namespaces, performance tuning
Strong Python backend development (FastAPI, async, SQLAlchemy)
Comfortable building Python control-plane agents that talk to Kubernetes APIs
Modern frontend development (TypeScript, React/Next.js, Tailwind, shadcn) - enough to ship product surfaces end-to-end
REST and tRPC API design
Experience building developer tools, dashboards, and live-monitoring UIs
Cash compensation $150K–$300K with significant equity
Flexible work arrangement (remote or San Francisco office)
Full visa sponsorship and relocation support
Professional development budget for courses and conferences
Regular team off-sites and conference attendance
Opportunity to shape the future of decentralized AI development
You'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source work.
We value potential over perfection - if you're passionate about democratizing AI development and have experience in either platform or infrastructure development (ideally both), we want to talk to you.
Ready to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team. \n\n\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\n\n\nROLE IMPACT\n\nYou'll help build our hosted training platform - the product that lets users launch LoRA and full fine-tuning runs on managed GPU clusters with a single API call or a few clicks. The role spans the developer-facing platform and the underlying Kubernetes-based training infrastructure that runs the jobs. \n\n\n\n\nCORE TECHNICAL RESPONSIBILITIES \n\n\n\n\nHOSTED TRAINING INFRASTRUCTURE \n\n - Design and operate Kubernetes-based training and inference orchestration across multi-cluster, multi-cloud GPU fleets\n\n - Build and maintain Helm charts that compose trainers, inference servers, environment servers, and supporting services into reproducible \"Training stacks\"\n\n - Develop the Python control-plane agents that watch pods, report run state to the platform, and keep clusters in sync\n\n - Implement scheduling and autoscaling for heterogeneous hardware (H100/H200/B200) using KEDA, LeaderWorkerSet, taints/tolerations, and gang scheduling\n\n - Run a tight GitOps workflow - every change ships through PRs, Helm values, and CI\n\n - Build node-local model caches, checkpoint pipelines, and shared storage for fast cold starts\n\n - Operate the observability stack (Prometheus, Grafana, Loki, DCGM) and make GPU cluster debugging fast\n\n\nPLATFORM DEVELOPMENT\n\n - Build the developer-facing surfaces for hosted training: job submission, live run monitoring, logs, metrics, model/adapter management, comparisons\n\n - Develop FastAPI backend services and REST APIs that bridge the platform to running clusters\n\n - Build real-time monitoring and debugging tools (streaming logs, step-level metrics, failure analysis)\n\n - Ship product UI in Next.js / React / TypeScript with shadcn, Tailwind, tRPC, and TanStack Query\n\n\nRESEARCH BRIDGE\n\n - Interface with the RL trainer, inference servers, and environment servers running inside our clusters\n\n - Productize new training capabilities (new model architectures, RL algorithms, modes)\n\n\n\n\n\nTECHNICAL REQUIREMENTS\n\nWe're looking for engineers who are fluent across three areas - you don't need to be the world's best at any one, but you should have real depth in all three and a clear point of view on how they connect.\n\n\n\n\nAI & GPU LANDSCAPE\n\n - Strong working knowledge of the modern AI stack - open model families, finetuning techniques (LoRA, QLoRA, full FT, RLHF/RLAIF), inference engines (vLLM, SGLang, TensorRT-LLM)\n\n - Familiarity with GPU hardware tradeoffs (H100 / H200 / B200, NVLink, interconnects, memory hierarchy) and what they mean for training and inference workloads\n\n - Understanding of distributed training fundamentals (data/tensor/pipeline/expert parallelism, NCCL, multi-node scheduling)\n\n - Awareness of what's happening at the frontier - new models, training methods, infra patterns - and the ability to translate that into product decisions\n \n \n\n\nKUBERNETES & INFRASTRUCTURE\n\n - Strong Kubernetes operations experience - Helm, CRDs, operators, KEDA, gang scheduling, GPU operator\n\n - Comfortable debugging real production clusters (kubectl, pod lifecycle, node issues, networking)\n\n - Cloud platform experience (GCP preferred - GCS, GKE, Cloud Run, Cloud Tasks)\n\n - Infrastructure automation (Helm, Terraform, Ansible) and a GitOps mindset\n\n - Observability: Prometheus, Grafana, Loki, OpenTelemetry, DCGM\n\n - Linux fundamentals: networking, namespaces, performance tuning\n \n \n\n\nPROGRAMMING & PLATFORM\n\n - Strong Python backend development (FastAPI, async, SQLAlchemy)\n\n - Comfortable building Python control-plane agents that talk to Kubernetes APIs\n\n - Modern frontend development (TypeScript, React/Next.js, Tailwind, shadcn) - enough to ship product surfaces end-to-end\n\n - REST and tRPC API design\n\n - Experience building developer tools, dashboards, and live-monitoring UIs\n \n \n\n\nWHAT WE OFFER\n\n - Cash compensation $150K–$300K with significant equity\n\n - Flexible work arrangement (remote or San Francisco office)\n\n - Full visa sponsorship and relocation support\n\n - Professional development budget for courses and conferences\n\n - Regular team off-sites and conference attendance\n\n - Opportunity to shape the future of decentralized AI development\n \n \n\n\nGROWTH OPPORTUNITY\n\nYou'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source work.\n\nWe value potential over perfection - if you're passionate about democratizing AI development and have experience in either platform or infrastructure development (ideally both), we want to talk to you.\n\n\n\nReady to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone."},{"id":"82084989-3721-475b-8253-46faab74a0ab","title":"Compute Finance and Strategy","department":"Compute / Finance","team":"Compute / Finance","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-07-08T18:35:11.447+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/82084989-3721-475b-8253-46faab74a0ab","applyUrl":"https://jobs.ashbyhq.com/primeintellect/82084989-3721-475b-8253-46faab74a0ab/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Compute is the foundational input of the AI era. The companies, models, and capabilities that define the next decade will be shaped by who has access to compute, on what terms, at what economics, and how it gets allocated across the systems that get built on top of it. The financial and operational architecture for an asset class of this consequence is still being built, and the playbooks for navigating it don't yet exist. The people who write them will define how the AI infrastructure industry develops over the next decade.
You will own the analytical foundation for how we understand global compute markets: pricing supply across regions and term lengths, modeling the economics of large GPU commitments, evaluating neoclouds and hyperscalers, and turning that work into provider decisions, commercial structures, and customer-facing products.
The work sits at the intersection of infrastructure, finance, and AI systems. You will evaluate questions like when an H200 cluster is the right fit versus GB200 or GB300, how networking and storage constraints affect real workload performance, how utilization assumptions change the economics of a multi-year commitment, and how regional power, colo, and capital costs flow through to GPU-hour pricing. You will diligence providers not just on headline price, but on delivery timeline, cluster architecture, reliability, support model, contractual risk, and ability to serve frontier AI workloads.
The decisions you support will directly shape Prime Intellect’s ability to deliver high-quality compute to researchers, AI labs, and enterprises building on top of our stack.
Compute Economics
Build and own the financial models that price our compute supply: per-cluster economics, contract structures, hardware generation comparisons, geographic and provider differentials
Model the economics of every meaningful supply decision — reserved vs. spot tradeoffs, term length, commitment level, hardware generation, provider mix, geography
Own margin architecture: margin by workload, customer, product, and contract, so we always know what's actually profitable and where the leverage is
Model the long-term P&L consequences of today's supply bets under multiple demand and pricing scenarios
Strategic Bets & Capital Allocation
Partner with leadership on the biggest decisions the company makes: which providers to commit to, which hardware generations, what geographies to lean into, how aggressively to scale
Build the financial frameworks that turn ambiguous strategic questions into decisions we can make with conviction
Own the long-range plan, scenario models, and capital allocation framework across compute, headcount, and product investment
Provider Engagement & Diligence
Engage directly with neoclouds, hyperscalers, and emerging providers on economic and technical diligence
Run the financial side of supply qualification — what we accept, what we reject, what we negotiate harder on
Translate technical performance characteristics into commercial recommendations
Build the repeatable analytical process for evaluating new entrants to the global supply market
Market Intelligence
Track pricing, availability, and provider dynamics continuously across every major market
Build Prime Intellect's view of the global compute market — who's credible, who's mispriced, where supply is tightening, where the next wave of capacity is coming online
Develop the analytical basis for our market positioning: when to commit hard, when to hold flexibility, where to lean in geographically
Cross-functional Partnership
Partner with Strategic Finance on how compute economics flow through to the company P&L
Partner with Engineering on the technical performance characteristics that drive cluster economics
Partner with Sales and Product on pricing strategy for consumption-based and hybrid products
Build board-ready analyses on supply strategy, capital allocation, and market positioning
What We're Looking For
4–7+ years in roles that combine financial rigor with real-world strategic or operational engagement. Backgrounds we'd find compelling include:
Investment banking, private equity, or growth equity with exposure to infrastructure, cloud, semiconductors, or technology
Quantitative or strategist roles at hedge funds, commodities desks, or trading firms
Infrastructure investing, project finance, or structured credit
Strategic finance or BizOps at a high-growth cloud, AI infrastructure, or compute-intensive company
Exceptional modeling and analytical skills — you build the models yourself, and your models reflect how the business actually works
Genuine technical curiosity. You don't need deep technical background to start, but you should be excited to develop fluency in GPU architectures, networking, cluster performance, and what makes one piece of compute economically different from another
Strong commercial and strategic judgment — you understand that finance's job is to drive better decisions, not produce more analysis
Comfortable engaging directly with vendors, partners, and senior counterparts at provider companies
Ability to operate across registers — building rigorous models, briefing leadership on strategic implications, and running diligence with senior counterparts at provider companies
High ownership — you see gaps and build the fix before anyone asks
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Direct experience modeling datacenter, colocation, cloud, or power/energy economics
Background covering AI infrastructure, cloud providers, semiconductors, or compute marketplaces from the banking, investing, or trading side
Hands-on experience with cluster benchmarking, training/inference workload economics, or compute marketplaces
Compute economics is becoming one of the most consequential domains in technology, and almost no one is approaching it with the rigor it deserves. You'll be in the room for the decisions shaping Prime Intellect's future and, in real ways, the future of open AI infrastructure. You'll work directly with leadership on the calls that define the company, develop deep expertise in a market most finance professionals only read about, and build a foundation in compute economics that is increasingly valuable across the industry.
Cash Compensation Range of $200-300k + meaningful equity
Flexible work (remote or San Francisco)
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
As Head of Marketing & Developer Relations, you'll own how Prime Intellect shows up in the world — the story we tell, the community we build, and how developers and researchers discover, adopt, and champion our platform. You'll shape our brand, run high-signal campaigns and launches, grow our developer community, and build the content engine that makes Prime Intellect the default starting point for RL infrastructure. This is a high-leverage role combining narrative craft, technical credibility, and community instinct — ideal for someone who can move fluidly between writing a launch post, scoping a conference strategy, and shipping a developer tutorial.
Define and evolve Prime Intellect's brand positioning, voice, and visual identity across all channels.
Translate deeply technical capabilities (async RL, distributed training, agentic post-training) into clear, compelling market narratives.
Own messaging frameworks for product launches, fundraise announcements, research releases, and partnership news.
Build and lead the DevRel function — creating programs, content, and experiences that drive developer adoption and loyalty.
Grow and engage our open-source community (Environments Hub, prime-rl, Verifiers) through tutorials, office hours, hackathons, and contributor programs.
Develop and maintain developer documentation, quickstarts, and onboarding flows in close partnership with Product and Engineering.
Represent Prime Intellect at conferences, meetups, and on podcasts — and identify opportunities for the founding team to do the same.
Build a content engine spanning technical blog posts, research digests, case studies, video content, and social media.
Partner with Research to translate frontier work (Recursive Language Models, INTELLECT model series, CLaaS) into high-signal public content.
Develop a thought leadership program that positions Prime Intellect's founders and researchers as key voices in open AI infrastructure.
Plan and execute product launches, research releases, and community events that generate developer awareness and adoption.
Run growth experiments across channels — organic social, paid, SEO, email, partnerships — and build measurement systems to track what works.
Own top-of-funnel metrics: developer signups, community engagement, content reach, and brand awareness.
Partner with Sales/GTM to ensure marketing-generated demand converts into pipeline and design-partner relationships.
Build and manage the marketing tech stack (analytics, CMS, email, social scheduling, attribution).
Establish reporting cadence for marketing KPIs and campaign performance.
Manage external agencies, freelancers, and contractors as needed for design, video, and content production.
5+ years in marketing, developer relations, or developer marketing at a B2B infrastructure, open-source, or developer-tools company.
Deep technical fluency — you can credibly discuss GPUs, RL training loops, model architectures, and API design with engineers and researchers.
Proven track record building developer communities and running high-impact product launches.
Strong writer and storyteller who can produce or direct content that resonates with both technical and business audiences.
Analytical and experiment-driven — comfortable with attribution models, funnel metrics, and growth experiments.
Experience building small, high-performing teams from scratch in fast-moving environments.
Hands-on experience with AI/ML infrastructure, cloud platforms, or open-source ML tooling.
Existing presence and credibility in the AI research or infrastructure community (blog, Twitter/X, conference talks).
Experience with open-source community building and contributor growth programs.
Early-stage operator mindset — bias for building playbooks, not inheriting them.
Cash Compensation Range of $150–300k + equity incentives
Flexible Work (remote or San Francisco)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
Opportunity to shape open AI infrastructure at Prime Intellect
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Security is the single highest-stakes function at Prime Intellect. Our customers — from frontier AI labs to enterprises — trust us with their most valuable assets: proprietary models, training data, and the compute that powers them. This role owns the security posture of everything we ship: the hosted RL training platform, distributed GPU infrastructure, liquid compute marketplace, and all customer-facing surfaces.
You'll be the first dedicated security hire and will define how we think about security as a company — from threat modeling and secure architecture to incident response and compliance. You'll work directly with engineering, research, and leadership to embed security into every layer of the stack, and you'll manage relationships with external penetration testers and security auditors to continuously validate our defenses.
Own threat modeling across our entire surface area: multi-tenant training infrastructure, sandboxed execution environments, API surfaces, and internal tooling
Design and implement zero-trust networking, identity, and access control across distributed GPU clusters and cloud infrastructure
Build secure-by-default patterns for our platform engineers — auth, secrets management, supply chain integrity, container hardening
Architect tenant isolation and data boundary enforcement for hosted RL training workloads (customers run arbitrary code in our environments)
Develop security frameworks specific to AI infrastructure: model weight protection, training data isolation, checkpoint integrity, gradient privacy
Secure the RL training loop end-to-end — from environment execution in sandboxes to reward signal verification and model artifact storage
Build detection and prevention for AI-specific attack vectors: prompt injection across agentic pipelines, model exfiltration, adversarial environment manipulation
Scope, manage, and run point on external penetration tests across our platform, hosted training infrastructure, and liquid compute layer
Build and maintain an internal red-teaming practice — automated and manual — targeting our most critical systems
Drive vulnerability management: triage, remediation SLAs, and root cause analysis
Build security monitoring and alerting across infrastructure (distributed clusters, Kubernetes, cloud) and application layers
Implement runtime security for containerized training workloads and sandboxed environments
Own incident response — build the playbooks, run the drills, lead the post-mortems
Design audit logging and forensic capability across all customer-facing systems
Drive SOC 2 Type II readiness and other compliance frameworks required by enterprise customers
Own the security narrative for customer-facing materials — questionnaires, architecture reviews, trust documentation
Partner with GTM to unblock enterprise deals that depend on security posture
5+ years in security engineering, infrastructure security, or offensive security roles — ideally at companies operating multi-tenant cloud or compute infrastructure
Deep experience with cloud security (GCP preferred), Kubernetes security, and container runtime hardening
Hands-on ability to read, write, and audit code in Python and Rust (or strong systems-level language)
Experience with network security in distributed systems — service mesh, mTLS, network segmentation across heterogeneous hardware
Proven track record managing external penetration tests and translating findings into engineering action
Strong fundamentals in cryptography, identity/access management, and secure software development lifecycle
Experience securing GPU infrastructure or ML training pipelines
Background in offensive security — CTFs, bug bounties, red team engagements
Familiarity with AI-specific threat models (model extraction, training data poisoning, sandbox escape)
Experience building security programs from scratch at a high-growth startup
SOC 2, ISO 27001, or FedRAMP compliance experience
Open-source security tooling contributions
Familiarity with eBPF, Falco, or similar runtime security tools
Cash Compensation Range of $180-350k+ with significant equity incentives
Based in our San Francisco office.
Full visa sponsorship and relocation support
Professional development budget for courses and conferences
Regular team off-sites and conference attendance
Opportunity to shape the future of decentralized AI development
You'll be the foundational security hire at a company where security is existential — not a checkbox. As we scale to support frontier AI labs and enterprises running their most critical training workloads on our platform, you'll build the security org from the ground up and have direct influence on company strategy.
We value intensity and ownership over credentials. If you've built security programs at infrastructure companies, broken into systems professionally, or have deep conviction about how AI infrastructure should be secured — we want to talk.
Ready to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\n\n\nROLE IMPACT\n\nSecurity is the single highest-stakes function at Prime Intellect. Our customers — from frontier AI labs to enterprises — trust us with their most valuable assets: proprietary models, training data, and the compute that powers them. This role owns the security posture of everything we ship: the hosted RL training platform, distributed GPU infrastructure, liquid compute marketplace, and all customer-facing surfaces.\n\nYou'll be the first dedicated security hire and will define how we think about security as a company — from threat modeling and secure architecture to incident response and compliance. You'll work directly with engineering, research, and leadership to embed security into every layer of the stack, and you'll manage relationships with external penetration testers and security auditors to continuously validate our defenses.\n\n \n\n\nCORE TECHNICAL RESPONSIBILITIES\n\n\nPREVENTIVE SECURITY & SECURE ARCHITECTURE\n\n - Own threat modeling across our entire surface area: multi-tenant training infrastructure, sandboxed execution environments, API surfaces, and internal tooling\n\n - Design and implement zero-trust networking, identity, and access control across distributed GPU clusters and cloud infrastructure\n\n - Build secure-by-default patterns for our platform engineers — auth, secrets management, supply chain integrity, container hardening\n\n - Architect tenant isolation and data boundary enforcement for hosted RL training workloads (customers run arbitrary code in our environments)\n\n\nAI-NATIVE SECURITY\n\n - Develop security frameworks specific to AI infrastructure: model weight protection, training data isolation, checkpoint integrity, gradient privacy\n\n - Secure the RL training loop end-to-end — from environment execution in sandboxes to reward signal verification and model artifact storage\n\n - Build detection and prevention for AI-specific attack vectors: prompt injection across agentic pipelines, model exfiltration, adversarial environment manipulation\n\n\nOFFENSIVE SECURITY & EXTERNAL ENGAGEMENTS\n\n - Scope, manage, and run point on external penetration tests across our platform, hosted training infrastructure, and liquid compute layer\n\n - Build and maintain an internal red-teaming practice — automated and manual — targeting our most critical systems\n\n - Drive vulnerability management: triage, remediation SLAs, and root cause analysis\n\n\nDETECTION, RESPONSE & OBSERVABILITY\n\n - Build security monitoring and alerting across infrastructure (distributed clusters, Kubernetes, cloud) and application layers\n\n - Implement runtime security for containerized training workloads and sandboxed environments\n\n - Own incident response — build the playbooks, run the drills, lead the post-mortems\n\n - Design audit logging and forensic capability across all customer-facing systems\n\n\nCOMPLIANCE & CUSTOMER TRUST\n\n - Drive SOC 2 Type II readiness and other compliance frameworks required by enterprise customers\n\n - Own the security narrative for customer-facing materials — questionnaires, architecture reviews, trust documentation\n\n - Partner with GTM to unblock enterprise deals that depend on security posture\n\n\nTECHNICAL REQUIREMENTS\n\n - 5+ years in security engineering, infrastructure security, or offensive security roles — ideally at companies operating multi-tenant cloud or compute infrastructure\n\n - Deep experience with cloud security (GCP preferred), Kubernetes security, and container runtime hardening\n\n - Hands-on ability to read, write, and audit code in Python and Rust (or strong systems-level language)\n\n - Experience with network security in distributed systems — service mesh, mTLS, network segmentation across heterogeneous hardware\n\n - Proven track record managing external penetration tests and translating findings into engineering action\n\n - Strong fundamentals in cryptography, identity/access management, and secure software development lifecycle\n\n\nNICE TO HAVE\n\n - Experience securing GPU infrastructure or ML training pipelines\n\n - Background in offensive security — CTFs, bug bounties, red team engagements\n\n - Familiarity with AI-specific threat models (model extraction, training data poisoning, sandbox escape)\n\n - Experience building security programs from scratch at a high-growth startup\n\n - SOC 2, ISO 27001, or FedRAMP compliance experience\n\n - Open-source security tooling contributions\n\n - Familiarity with eBPF, Falco, or similar runtime security tools\n\n\nWHAT WE OFFER\n\n - Cash Compensation Range of $180-350k+ with significant equity incentives\n\n - Based in our San Francisco office.\n\n - Full visa sponsorship and relocation support\n\n - Professional development budget for courses and conferences\n\n - Regular team off-sites and conference attendance\n\n - Opportunity to shape the future of decentralized AI development\n\n\nGROWTH OPPORTUNITY\n\nYou'll be the foundational security hire at a company where security is existential — not a checkbox. As we scale to support frontier AI labs and enterprises running their most critical training workloads on our platform, you'll build the security org from the ground up and have direct influence on company strategy.\n\nWe value intensity and ownership over credentials. If you've built security programs at infrastructure companies, broken into systems professionally, or have deep conviction about how AI infrastructure should be secured — we want to talk.\n\nReady to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone."},{"id":"d523340c-9914-4b86-828a-147c9ada94d3","title":"Solutions Architect - AI Infrastructure","department":"Compute / Finance","team":"Compute / Finance","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-07-08T18:36:17.454+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/d523340c-9914-4b86-828a-147c9ada94d3","applyUrl":"https://jobs.ashbyhq.com/primeintellect/d523340c-9914-4b86-828a-147c9ada94d3/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
This role sits at the intersection of customer success, technical operations, and AI infrastructure delivery. You will work directly with customers to help them onboard, deploy, and scale on Prime Intellect infrastructure, while also building the internal systems, workflows, and automation that make that experience scalable over time.
Beyond supporting individual accounts, this person will help define how Prime Intellect serves sophisticated technical customers: turning high-touch support and deployment work into repeatable processes, better tooling, and durable operational leverage.
Serve as a primary point of contact for customers running AI workloads on our infrastructure
Own customer onboarding and deployment coordination from post-sale through production readiness
Build trusted relationships with customer teams and proactively manage issues, risks, and escalations
Partner with engineering and operations to ensure customer needs are translated clearly into internal action
Build and improve the internal systems that support customer onboarding, deployment tracking, escalation management, and account health
Identify repetitive manual workflows and turn them into streamlined processes, tooling, or automation
Help define the operating cadence, support infrastructure, and service model for managing technical customer accounts at scale
Create structure where none exists today, improving both customer experience and internal efficiency as the business grows
Track billing, contract milestones, renewals, and other key account workflows
Maintain visibility into deployment timelines, infrastructure commitments, and customer-specific requirements
Coordinate across customers, finance, engineering, and leadership to keep accounts running smoothly end-to-end
2–5 years of experience in customer success, technical account management, operations, or a similar cross-functional role
Strong interest in AI, GPU infrastructure, cloud systems, or technical products serving sophisticated customers
Comfortable working directly with customers and internal technical teams
Highly organized, detail-oriented, and able to manage multiple workstreams at once
Strong written and verbal communication skills
High ownership and willingness to operate flexibly when customer needs require urgency
Energized by building processes from the ground up in a fast-moving environment
Experience with infrastructure, cloud, DevOps, or enterprise technical products
Familiarity with SLAs, support operations, billing, procurement, or contract administration
Exposure to AI/ML infrastructure, model training environments, or GPU-based products
Experience in startups or high-growth environments
$200-275k Cash compensation + equity incentives
Flexible work environment
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conference attendance
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will own compute at Prime Intellect - the sourcing, economics, contracting, and strategic positioning of the GPU supply that powers everything we train, serve, and sell.
This is one of the most consequential roles at the company, and one of the most consequential compute roles in the industry. Compute is our product, and compute is the constraint that shapes what the open AI ecosystem can do. Every frontier lab is fighting for the same GPUs; your job is to make sure the open ecosystem has a leading seat at the table.
The decisions you make will shape the industry. Which neoclouds get the capital and commitments to scale. Which geographies become meaningful compute hubs. Which hardware generations get broad access versus staying locked inside closed labs. Which open models are economically possible to train and serve. You'll co-design the compute layer of the open model ecosystem alongside our research and engineering teams — deciding, together, what we train, on what, where, and at what cost structure.
You'll work directly with leadership team and founders of the neoclouds reshaping global compute, and with the research and engineering teams pushing the frontier of open post-training. You'll structure the commercial relationships that define the next several years of AI infrastructure, get first access to the latest generations of accelerators as they come online, and build the financial and operational architecture that turns a fragmented global supply market into Prime Intellect's durable advantage.
You need to be as comfortable modeling the unit economics of a three-year GB200 commitment against a volatile spot market as you are sitting with our research team to understand what an upcoming training run actually needs, or negotiating a nine-figure reserved capacity agreement with neocloud leadership.
Compute Strategy & Economics
Own the economics of compute end-to-end: the unit economics of every contract, the margin architecture across training and inference products, the long-term P&L consequences of today's supply bets
Partner with Finance and leadership on capital strategy — how much to commit, to whom, for how long, on which hardware, with what balance sheet exposure
Build the frameworks that turn supply decisions into clear financial outcomes, and that let us make multi-hundred-million-dollar bets with conviction under uncertainty
Shape the commercial architecture of the open compute ecosystem: how committed capacity, spot markets, credit structures, and partner economics fit together
Sourcing & Contracting
Own end-to-end procurement of GPU capacity globally — across hyperscalers, tier-one neoclouds, regional operators, and emerging providers in North America, Europe, the Middle East, and Asia
Negotiate and close reserved capacity agreements, spot and burst arrangements, MSAs, DPAs, and order forms at nine- and ten-figure scale
Secure early access to the latest generations of accelerators (B200, GB200, and what comes next) — in the quantities we need, before our competitors
Build and maintain the senior relationships that make Prime Intellect the partner of choice for providers deciding where to allocate scarce capacity
Co-Design with Research & Engineering
Work closely with our research team to translate training roadmaps, RL workloads, and open model ambitions into concrete compute requirements — and back the other way, to surface what's possible given the supply we can secure
Partner with Engineering on acceptance testing, goodput validation, and the technical qualification of new providers and hardware
Sit at the table where the biggest calls get made: which open models we train, which customers we serve, which bets are worth the capital
Market Intelligence & Positioning
Track pricing, availability, and provider dynamics continuously across every major global market
Serve as the internal source of truth on the compute market — who's credible, who's mispriced, where supply is about to tighten, which providers will still exist in 18 months, where the next wave of capacity is coming online
Advise leadership on the strategic bets that define the company: which accelerators, which providers, which geographies, which contract structures, which moments to lean in hard
Strong business and financial instincts — you think natively in unit economics, margin structure, and capital allocation, and you can model the long-term P&L consequences of complex supply decisions
Deep fluency in the global AI compute market: you know the providers, the hardware generations and their real tradeoffs, the pricing dynamics, and where the market is going over the next 12–24 months
Enough technical understanding to be dangerous — you can push back on a vendor's spec sheet, read a cluster topology diagram, understand why two nominally identical clusters deliver different goodput, and have a real conversation with researchers about what their workloads need
Serious commercial chops: you've negotiated and closed contracts at meaningful scale, know how to find and use leverage, and understand how deal structure drives downstream economics
Comfortable operating at the intersection of finance, commercial, product, and engineering — and translating fluently between all of them
High ownership: you see gaps and build the fix before anyone asks
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Competitive Cash Compensation + meaningful equity
Flexible work (remote or San Francisco)
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You'll build the distributed control plane and node-level runtime behind our sandbox platform: placing workloads across bare-metal workers, launching secure microVMs, and making the entire lifecycle fast, efficient, and resilient to failure.
Build the control plane that places sandbox workloads efficiently across bare-metal CPU and GPU workers.
Build node-level execution systems that launch and supervise secure microVMs, coordinate storage and networking, and take sandboxes from assignment to ready execution.
Own the sandbox product surface, including its APIs, SDK, and CLI, and evolve it to support reinforcement learning and agentic workflows at scale.
Work closely with research, applied research, customers, and the broader engineering team to take problems from discovery through design, implementation, and production adoption.
You have 5+ years of experience building production distributed systems from the ground up.
You have strong systems programming experience. Production Go is a major advantage; Rust, C++, or comparable systems-language experience is also valued.
You are a high-agency engineer who can lead the design and implementation of high-performance, fault-tolerant systems.
You have strong Linux systems fundamentals and depth in at least one of virtualization, networking, storage, or workload isolation.
You enjoy going deep on systems performance, whether the bottleneck is memory behavior, concurrency, networking, or resource utilization.
You have built, deployed, and operated production systems at scale.
Cash compensation of $150-300k with significant equity incentives
Based in our San Francisco office.
Full visa sponsorship and relocation support
A professional development budget for courses and conferences
Regular team off-sites and conference attendance
The opportunity to shape core infrastructure for frontier AI research
You'll work alongside experienced engineers and researchers building core infrastructure for frontier AI research. You'll shape the sandbox platform that powers demanding reinforcement learning, evaluation, and agentic workloads, with opportunities to contribute to open-source infrastructure.
On a small, high-ownership team, you'll help set technical direction and engineering practices while moving quickly from decisions to shipped systems.
We value deep systems expertise and customer obsession. If you want to build foundational infrastructure from the control plane down to the node runtime, we want to talk to you.
Apply now and help build the systems behind the next generation of AI research.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n \n\n\nCORE TECHNICAL RESPONSIBILITIES\n\nYou'll build the distributed control plane and node-level runtime behind our sandbox platform: placing workloads across bare-metal workers, launching secure microVMs, and making the entire lifecycle fast, efficient, and resilient to failure.\n\n - Build the control plane that places sandbox workloads efficiently across bare-metal CPU and GPU workers.\n\n - Build node-level execution systems that launch and supervise secure microVMs, coordinate storage and networking, and take sandboxes from assignment to ready execution.\n\n - Own the sandbox product surface, including its APIs, SDK, and CLI, and evolve it to support reinforcement learning and agentic workflows at scale.\n\n - Work closely with research, applied research, customers, and the broader engineering team to take problems from discovery through design, implementation, and production adoption.\n\n \n\n\nTECHNICAL REQUIREMENTS\n\n - You have 5+ years of experience building production distributed systems from the ground up.\n\n - You have strong systems programming experience. Production Go is a major advantage; Rust, C++, or comparable systems-language experience is also valued.\n\n - You are a high-agency engineer who can lead the design and implementation of high-performance, fault-tolerant systems.\n\n - You have strong Linux systems fundamentals and depth in at least one of virtualization, networking, storage, or workload isolation.\n\n - You enjoy going deep on systems performance, whether the bottleneck is memory behavior, concurrency, networking, or resource utilization.\n\n - You have built, deployed, and operated production systems at scale.\n\n \n\n\nWHAT WE OFFER\n\n - Cash compensation of $150-300k with significant equity incentives\n\n - Based in our San Francisco office.\n\n - Full visa sponsorship and relocation support\n\n - A professional development budget for courses and conferences\n\n - Regular team off-sites and conference attendance\n\n - The opportunity to shape core infrastructure for frontier AI research\n\n \n\n\nGROWTH OPPORTUNITY\n\nYou'll work alongside experienced engineers and researchers building core infrastructure for frontier AI research. You'll shape the sandbox platform that powers demanding reinforcement learning, evaluation, and agentic workloads, with opportunities to contribute to open-source infrastructure.\n\nOn a small, high-ownership team, you'll help set technical direction and engineering practices while moving quickly from decisions to shipped systems.\n\nWe value deep systems expertise and customer obsession. If you want to build foundational infrastructure from the control plane down to the node runtime, we want to talk to you.\n\nApply now and help build the systems behind the next generation of AI research."},{"id":"d70f38c9-ae8e-4252-933a-666cb900e3b4","title":"Technical Account Manager - AI Infrastructure","department":"Compute / Finance","team":"Compute / Finance","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-07-08T18:36:01.490+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/d70f38c9-ae8e-4252-933a-666cb900e3b4","applyUrl":"https://jobs.ashbyhq.com/primeintellect/d70f38c9-ae8e-4252-933a-666cb900e3b4/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Your Role
Prime Intellect serves some of the most sophisticated AI teams in the world that depend on our compute and infrastructure to train and deploy production AI systems. The Customer Success Manager is the person who makes sure those customers succeed, scale, and keep building with us.
This is not a traditional Customer Success role. Our customers run large-scale training jobs, scale inference workloads against real production traffic, and depend on cluster reliability and performance the way most companies depend on their cloud provider. The work spans the technical and the commercial — you'll be reading Grafana dashboards and discussing cluster performance with a customer's ML infrastructure team in the morning, and partnering with Sales on a capacity expansion in the afternoon.
You'll own a portfolio of enterprise customers end-to-end and build the relationships that make Prime Intellect the partner of choice for their AI infrastructure.
Responsibilities
Customer Ownership
Own a portfolio of enterprise customers end-to-end — adoption, retention, expansion, and overall health
Build deep relationships with technical and executive stakeholders at each customer, from ML engineers to engineering leadership
Drive customer outcomes: faster time-to-value on first workloads, smooth scaling as their usage grows, and meaningful expansion as their AI ambitions expand
Technical Partnership
Understand each customer's training and inference workloads at a real technical level — what models they're training, what infrastructure they need, what their performance bottlenecks are
Partner with customers' engineering teams on cluster performance, capacity planning, workload optimization, and migration
Translate customer needs into clear, prioritized feedback for our Engineering and Product teams
Expansion & Renewals
Identify expansion opportunities ahead of the customer — anticipate scaling needs, surface new use cases, drive adoption of new products (Lab, Inference, additional compute capacity)
Partner with Sales on renewal conversations and growth motions
Maintain visibility into the economics of each customer relationship, in partnership with Finance and Compute
Operational Excellence
Serve as the first line for customer-facing operational issues — usage questions, capacity changes, SLA tracking, incident communications
Build the cross-functional connective tissue between Sales, Engineering, Finance, and customers
What We're Looking For
3–6 years in Customer Success, Technical Account Management, Solutions Engineering, or adjacent roles at infrastructure, cloud, or AI/ML companies
Strong technical fluency — comfortable reading dashboards, discussing infrastructure architecture, and engaging with customer engineering teams without a translator
Strong commercial instincts — you understand that Customer Success is a revenue function, not a support function, and you can drive real expansion alongside technical outcomes
Deep customer empathy combined with high judgment — you advocate for customers internally while making the calls that are right for the business
Excellent verbal and written communication, especially when explaining complex technical issues to non-technical stakeholders and vice versa
High ownership — you see gaps and build the fix before anyone asks
Comfortable in ambiguity and speed; this market doesn't slow down
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Direct experience at a cloud provider, AI infrastructure company, or compute marketplace
Familiarity with GPU economics, training and inference workloads, or compute consumption patterns
Background as a TAM or Solutions Architect at a hyperscaler (AWS, GCP, Azure) or specialized cloud provider
Working knowledge of usage-based pricing, capacity commitments, and consumption-based contracts
You've been an early Customer Success hire at a high-growth company
What We Offer
Cash Compensation Range of $160,000 – $200,000 + meaningful equity
Flexible work (remote or San Francisco)
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Every serious AI team evaluating infrastructure asks the same question: will this actually run my workload better, faster, or cheaper than what I have today? Most vendors answer with slides. We want to answer with benchmarks on the customer's own workload, and then design the production architecture that delivers on them.
As our Founding Solutions Architect, your mission is to make Prime Intellect the technical winner in every deal. You will own the path from workload discovery through architecture, benchmark, proof of concept, and production design. You will sit with customers' ML engineers to understand what they are actually training and serving — GPU type, cluster requirements, networking, utilization and performance objectives — then design the right architecture on Prime infrastructure, run the POC against their baseline and the competitive alternatives, and turn latency, throughput, utilization, and cost results into a technical business case that holds up under scrutiny.
This is a founding role. There is no playbook yet for how we prove technical wins; you will write it. The benchmark methodologies, reference architectures, and POC frameworks you build will become how Prime Intellect sells to every lab and enterprise that follows.
Technical Discovery & Architecture
Lead deep technical discovery across training, post-training, and inference workloads
Understand each customer's GPU requirements, cluster topology, networking, storage, utilization patterns, and performance objectives
Design customer architectures on Prime infrastructure that meet those objectives and are ready for production
Partner with engineering on migration and deployment requirements so what we design is what gets shipped
Benchmarks, POCs & the Technical Business Case
Own workload benchmarks and proofs of concept against customer baselines and competitive alternatives
Translate latency, throughput, utilization, and cost results into a defensible technical business case for ML engineers and executive buyers alike
Set the standard for benchmark quality: methodology, reproducibility, and honest comparison
Drive time to technical validation down with every deal
Playbooks & Platform Adoption
Develop repeatable benchmark and reference-architecture playbooks that scale beyond you
Support technical evaluation and adoption of Model Factory, bringing customers from compute onto the full training stack
Feed what you learn from customer workloads back into product and research priorities
Strong ML infrastructure, distributed systems, cloud, or GPU-compute background
Working knowledge of modern training and inference stacks, CUDA/GPU environments, and model-serving patterns
Customer-facing experience in solutions architecture, sales engineering, forward-deployed engineering, or a similar role
Ability to communicate fluently with both ML engineers and executive stakeholders — and to know which conversation you are in
A strong bias toward proving claims with real customer workloads and benchmarks rather than slides
High ownership — you see gaps and build the fix before anyone asks
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Hands-on experience running or optimizing large-scale distributed training (FSDP, Megatron, DeepSpeed, or similar) or high-throughput inference (vLLM, SGLang, TensorRT-LLM)
Experience benchmarking across hardware generations (H100, H200, GB200) and interconnects (InfiniBand, RoCE)
Background at a GPU cloud, neocloud, hyperscaler AI team, or frontier lab infrastructure team
Experience building a solutions or forward-deployed function from scratch
Technical win rate · POC-to-commercial conversion · Time to technical validation · Benchmark quality · Deployment readiness · Model Factory technical attach
You will be the technical face of Prime Intellect in every deal that matters, working directly with the ML teams at labs and enterprises deciding where to build — and with our own research and engineering teams on what the platform needs to win. You will see more real-world training and inference workloads in a year than most engineers see in a career, and you will define how a new category of infrastructure proves itself.
Cash compensation range of $250–$320k + meaningful equity
Flexible work (remote or San Francisco), with travel to customer sites as needed
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Compute is the scarcest input in AI, and the companies that secure it on the right terms will define the next decade. But the labs and enterprises we work with are not buying GPU hours. They are buying the ability to execute an AI roadmap: train the next model, post-train it on their own data, serve it in production, and do it again. Selling to them means understanding that roadmap and structuring a multi-year commitment around it.
As a Strategic Account Executive, you will own strategic customer acquisition and expansion for Prime Intellect's GPU infrastructure and Model Factory platform. You will build and close a portfolio of AI-native and enterprise accounts, working with CTOs, Heads of AI/ML, infrastructure leaders, and economic buyers to translate their training, post-training, and inference plans into committed capacity. You will construct deals around spend, term, workload expansion, and platform adoption — not feature checklists — and then grow every landed customer from compute into the full training stack.
You will work alongside our Solutions Architects on discovery, benchmarks, and POCs, and directly with leadership on the commercial structures that define how this category gets sold. The deals you close will shape what Prime Intellect commits to, builds, and becomes.
Strategic Acquisition
Build and close a portfolio of strategic AI-native and enterprise accounts
Develop relationships with CTOs, Heads of AI/ML, infrastructure leaders, and economic buyers
Translate customers' training, post-training, and inference roadmaps into multi-year GPU capacity commitments
Partner with Solutions Architects on technical discovery, benchmarks, and POCs so every deal is won on proof
Deal Construction & Expansion
Build commercial proposals around committed spend, term length, workload expansion, and Model Factory adoption
Negotiate and close seven- and eight-figure agreements with economics that work for both sides
Expand landed compute customers into Model Factory, post-training, inference, and additional capacity
Partner with Compute and Finance on supply, pricing, and margin so what you sell is what we can deliver
Pipeline Discipline
Maintain rigorous qualification, forecasting, and next-step discipline across the portfolio
Report pipeline, forecast, and deal status to leadership with accuracy you stand behind
Feed market, competitive, and customer signal back into product, pricing, and compute strategy
Proven success selling highly technical infrastructure: cloud, GPU compute, AI/ML platforms, or adjacent developer infrastructure
Experience closing seven-figure enterprise agreements; eight-figure strongly preferred
Credibility with technical executives and ML/infrastructure teams — you can hold the room with a Head of AI and an engineer at the same time
Ability to sell economics, architecture, and strategic partnership, not just software features
Strong deal construction and negotiation skills, including multi-year commitments and consumption-based structures
Comfort operating in an early-stage, high-accountability environment where you own the number
High ownership — you see gaps and build the fix before anyone asks
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Experience selling GPU capacity, reserved compute, or AI training infrastructure specifically
Background at a GPU cloud, neocloud, hyperscaler AI organization, or AI platform company
Existing relationships with AI labs, AI-native startups, or enterprise AI leaders
Experience as an early sales hire who helped define the playbook, pricing, and segmentation
Committed contract value · Contract duration · New strategic customers · GPU capacity sold · Model Factory attach · Post-training and inference expansion · Forecast accuracy
You will sell one of the most consequential products in technology to the people building the next generation of AI, in a market where the commercial playbook is still being written. You will work directly with leadership on deal structures that define the category, develop a network across the labs and enterprises shaping the industry, and build a track record in AI infrastructure sales that is increasingly rare and valuable.
On-target earnings of $225–$300k++ and meaningful equity
Flexible work (remote or San Francisco), with travel to customers as needed
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Your Role
Compute is the foundational input of everything Prime Intellect does — and right now, the picture of our compute supply, demand, and economics lives across spreadsheets, partner conversations, and people's heads. This role changes that.
As Compute Intelligence Engineer, you'll build the data infrastructure and intelligence platform that gives the entire company a live, accurate picture of our compute: what we have, what's coming online, where our bottlenecks are, and how supply maps to demand. This is a hands-on data engineering build — you'll stand up the warehouse, write the pipelines that pull from our compute telemetry, billing systems, partner data, and CRM, model that data into a clean and trustworthy source of truth, and turn it into dashboards and a queryable layer the whole company relies on.
This is a builder-first role with a clear business purpose. You won't be building data infrastructure for its own sake — you'll be building the system that lets our Compute Partnerships team, Growth team, and Research team operate from the same source of truth. When Growth needs to know what capacity is coming online next quarter, when Compute Partnerships needs to understand our utilization against commitments, when Research needs to scale a training run — the platform you build is what they'll turn to.
You'll be early in this seat, and the foundations you lay will be the data backbone the company scales on.
Responsibilities
Build the Compute Intelligence Platform
Stand up Prime Intellect's data warehouse (Snowflake, BigQuery, or equivalent) and the pipelines that feed it — compute telemetry, billing and usage data, partner and supply data, CRM, and financial systems
Build the data models and transformations (dbt or equivalent) that turn raw data into a clean, queryable, trustworthy source of truth
Build dashboards and reporting that give the company a live picture of compute supply, demand, utilization, upcoming capacity, and bottlenecks
Build a queryable, AI-accessible layer on top of the warehouse so teams across the company can answer their own questions without going through a data analyst
Supply & Demand Intelligence
Build the data systems that track our compute supply end-to-end: what we have, what's committed, what's coming online, and what's utilized vs. idle
Develop the views and models that surface where our bottlenecks are — and make upcoming supply legible to the teams that depend on it
Connect supply data to demand signals so the company can see, in one place, how capacity maps to what we're selling and building
Cross-Functional Enablement
Serve as the data backbone connecting Compute Partnerships, Growth, and Research — building the systems that let them operate from shared, accurate information
Partner with Growth on understanding upcoming supply and how it maps to what they can sell
Partner with Compute Partnerships on utilization, commitments, and supply tracking
Partner with Research on scaling needs and capacity planning
Operational Reliability
Build pipelines and systems that run unattended, stay in sync, and fail gracefully
Establish the data quality, documentation, and infrastructure standards that let the data layer scale with the company
Partner with Engineering on shared infrastructure, security, and data standards
What We're Looking For
3–7+ years in data engineering, analytics engineering, GTM/growth engineering, or similar roles where you've built data infrastructure that served real business outcomes
Strong technical skills: comfortable building and maintaining data warehouses, writing production-quality pipelines (Python, SQL), modeling data (dbt or equivalent), and connecting disparate systems via APIs
Experience with modern data stack tooling — Snowflake / BigQuery / Databricks, dbt, orchestration (Airflow, Dagster, etc.), and BI/dashboarding tools
A builder's instinct paired with business judgment — you don't just build what's asked; you understand the business well enough to build the right thing
Comfortable being the data backbone for cross-functional teams — translating between business needs and the systems that serve them
Familiarity with modern AI tooling and an interest in building AI-accessible data layers (natural-language querying, LLM-powered analytics) that let non-technical teams self-serve
High ownership — you see gaps and build the fix before anyone asks
Comfortable in ambiguity and speed; you'll be defining what the data layer looks like from scratch
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Experience as an early data hire who built a company's data infrastructure from scratch
Familiarity with GPU economics, compute infrastructure, cloud telemetry, or AI/ML workloads
Background in GTM engineering, growth engineering, or revenue/operations data
Experience building LLM-powered or natural-language data interfaces
Working knowledge of usage-based / consumption-based business models and the data they generate
What We Offer
Cash Compensation Range of $225-300k + meaningful equity
Flexible work (remote or San Francisco)
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more.
Talent is the single highest-leverage input at Prime Intellect. Every frontier researcher, systems engineer, and operator we hire compounds directly into what the open ecosystem can build. This role owns that compounding.
As Head of Talent, you will own recruiting end-to-end as we scale from ~40 to ~100 people over the next twelve months, weighted heavily toward frontier research and infrastructure engineering — the two most competitive talent markets in tech. You will personally run our most important searches, source candidates who are not on the market, and close them against offers from the largest labs.
Beyond individual hires, you will define how Prime Intellect recruits at scale: building the sourcing engine, closing playbooks, interview architecture, and compensation frameworks that let us maintain an exceptionally high bar while moving faster than incumbents. This is a founding recruiting leader role reporting directly to the founders, with full ownership of the function, meaningful equity, and a direct line into every strategic hiring decision.
Own the Highest-Priority Searches
Personally run our most important searches across research, infrastructure, and leadership hires
Source candidates directly — through your own network, cold outreach, and targeted campaigns — for roles where inbound alone will never surface the right person
Close candidates against competing offers from frontier labs, hyperscalers, and top AI startups
Serve as a trusted partner to the founders and hiring managers on candidate strategy, leveling, and offer construction
Build the Sourcing & Closing Engine
Design and run the sourcing motion for frontier research talent (labs, PhD programs, top open-source contributors) and elite systems engineering talent (distributed systems, GPU infrastructure, ML systems)
Build closing playbooks that reflect what actually moves technical candidates: research autonomy, compute access, publication norms, equity, and mission
Stand up the tooling, data, and pipeline visibility we need to run 60+ hires per year without dropping quality
Define and enforce the hiring bar across the company as we scale
Systems, Team & Operational Scale
Hire and lead a small recruiting team over time (sourcers, coordinators, and eventually recruiters), staying deeply hands-on yourself
Build interview architecture — rubrics, calibration, debriefs, decision-making — that produces consistent hiring outcomes as more people participate
Own recruiting operations, candidate experience, and reporting end-to-end
Partner with finance and the founders on compensation philosophy, leveling, and equity strategy as we grow
5+ years of technical recruiting experience, with meaningful time at a frontier AI lab, top AI startup, or elite technical company
Direct experience recruiting ML researchers, frontier research talent, or elite systems / distributed engineers — not just \"technical roles\" broadly
Proven ability to source and close candidates who are not actively looking, against competing offers from top labs
Track record of personally owning senior and hard-to-fill searches end-to-end
Comfort operating as a player-coach: running your own searches while building the systems and eventually the team underneath you
Strong judgment on candidate quality and leveling; willingness to hold a high bar under pressure
High ownership, direct communication, and appetite for working closely with founders on the most consequential decisions in the company
Energized by building a recruiting function from the ground up in a fast-moving environment
Prior experience as a founding recruiter, Head of Recruiting, or senior recruiting lead at a fast-growing technical company
Existing network in the frontier AI research or GPU / ML infrastructure community
Experience closing candidates against top AI lab offers
Familiarity with open-source ecosystems and the researchers and engineers who lead them
Background scaling recruiting operations, tooling, and interview systems at a rapidly growing company
$250-300k+ cash compensation + meaningful equity
Flexible work environment
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conference attendance
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will work on reinforcement learning for agents that reason over long horizons, use tools, coordinate with other agents, recover from failure, and improve from experience. This is Prime's broad, cross-domain Applied Research role, focused on RL and ML systems. Your work should hold up across several application domains: algorithms, environments, harnesses, reward and credit-assignment methods, evaluation, and scalable experimental systems. You will work regularly with application teams, partners, and open-source users while remaining focused on methods and systems that transfer across domains.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
Recent Prime work includes verifiers v1, an algorithms layer for prime-rl, multi-agent training, Prime Agent, and RL systems tested at trillion-parameter scale. You will extend this line of work and pressure-test it against live application programs.
Representative early problems
Build agentic judges and credit-assignment methods that inspect complete trajectories, use tools, and assign useful learning signal beyond one final score.
Make multi-agent and long-horizon environments train reliably at scale while preserving debuggable traces and comparable baselines.
Detect when apparent gains come from reward hacking, harness behavior, data leakage, or infrastructure artifacts, then redesign the loop so the result survives.
What you will own
RL and post-training methods for tool-using, multi-agent, memory-augmented, and long-horizon systems.
Reusable environments, harnesses, rewards, and evaluations.
Experiments that connect algorithm choices to training and inference behavior.
Open-source systems and papers built from problems surfaced by application researchers and external users.
Relevant experience: hands-on RL training, scalable ML systems, agent environments, evaluation, inference/training infrastructure, or research crossing methods and systems.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou will work on reinforcement learning for agents that reason over long horizons, use tools, coordinate with other agents, recover from failure, and improve from experience. This is Prime's broad, cross-domain Applied Research role, focused on RL and ML systems. Your work should hold up across several application domains: algorithms, environments, harnesses, reward and credit-assignment methods, evaluation, and scalable experimental systems. You will work regularly with application teams, partners, and open-source users while remaining focused on methods and systems that transfer across domains.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nRecent Prime work includes verifiers v1 https://www.primeintellect.ai/blog/verifiers-v1, an algorithms layer for prime-rl https://www.primeintellect.ai/blog/algorithms-layer, multi-agent training https://www.primeintellect.ai/blog/multi-agent-systems, Prime Agent https://www.primeintellect.ai/blog/prime-agent, and RL systems tested at trillion-parameter scale https://www.primeintellect.ai/blog/rl-at-1t-scale. You will extend this line of work and pressure-test it against live application programs.\n\nRepresentative early problems\n\n - Build agentic judges and credit-assignment methods that inspect complete trajectories, use tools, and assign useful learning signal beyond one final score.\n\n - Make multi-agent and long-horizon environments train reliably at scale while preserving debuggable traces and comparable baselines.\n\n - Detect when apparent gains come from reward hacking, harness behavior, data leakage, or infrastructure artifacts, then redesign the loop so the result survives.\n\nWhat you will own\n\n - RL and post-training methods for tool-using, multi-agent, memory-augmented, and long-horizon systems.\n\n - Reusable environments, harnesses, rewards, and evaluations.\n\n - Experiments that connect algorithm choices to training and inference behavior.\n\n - Open-source systems and papers built from problems surfaced by application researchers and external users.\n\n\nREQUIREMENTS\n\n - Relevant experience: hands-on RL training, scalable ML systems, agent environments, evaluation, inference/training infrastructure, or research crossing methods and systems.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"af2f419f-6596-4805-8e83-9cee5016aa1a","title":"Applied Research - Forward-Deployed RL","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:29:13.997+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/af2f419f-6596-4805-8e83-9cee5016aa1a","applyUrl":"https://jobs.ashbyhq.com/primeintellect/af2f419f-6596-4805-8e83-9cee5016aa1a/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will own applied RL programs from the first ambiguous customer problem through a deployed improvement loop. You will understand the workflow and business stakes, decide what can be measured and learned, build the environment and evaluation, run post-training experiments, deploy the result, and determine what should generalize into Prime's shared stack. The majority of your work will be organized around external programs. You will work directly with customer researchers, engineers, and leaders, and you will be accountable for both the research result and its practical impact.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
We have active programs across technology, science, engineering, regulated industries, and the public sector.
Representative early problems
Turn a partially documented workflow and business goal into an environment, baseline, and evaluation that the customer team agrees measures the right outcome.
Improve a domain agent through RL or another post-training method while separating model limitations from task, harness, inference, and integration failures.
Carry a successful experiment into deployment, measure its effect on the customer outcome, and convert the recurring parts into reusable environments, recipes, data loops, or platform capabilities.
What you will own
Technical and business discovery that produces a clear research plan, success measure, environment, and baseline.
RL, distillation, and other post-training experiments tied to deployment outcomes.
Agent systems that handle tool use, memory, recovery, long horizons, and human oversight.
Deployment and adoption decisions, plus reusable artifacts carried back to Prime's research, engineering, and product teams.
Relevant experience: customer-embedded technical work, research engineering, technical founding experience, applied ML consulting with real build ownership, or end-to-end post-training programs.
You should be comfortable with senior external teams and able to connect experimental evidence to business decisions without lowering the research bar.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou will own applied RL programs from the first ambiguous customer problem through a deployed improvement loop. You will understand the workflow and business stakes, decide what can be measured and learned, build the environment and evaluation, run post-training experiments, deploy the result, and determine what should generalize into Prime's shared stack. The majority of your work will be organized around external programs. You will work directly with customer researchers, engineers, and leaders, and you will be accountable for both the research result and its practical impact.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nWe have active programs across technology, science, engineering, regulated industries, and the public sector.\n\nRepresentative early problems\n\n - Turn a partially documented workflow and business goal into an environment, baseline, and evaluation that the customer team agrees measures the right outcome.\n\n - Improve a domain agent through RL or another post-training method while separating model limitations from task, harness, inference, and integration failures.\n\n - Carry a successful experiment into deployment, measure its effect on the customer outcome, and convert the recurring parts into reusable environments, recipes, data loops, or platform capabilities.\n\nWhat you will own\n\n - Technical and business discovery that produces a clear research plan, success measure, environment, and baseline.\n\n - RL, distillation, and other post-training experiments tied to deployment outcomes.\n\n - Agent systems that handle tool use, memory, recovery, long horizons, and human oversight.\n\n - Deployment and adoption decisions, plus reusable artifacts carried back to Prime's research, engineering, and product teams.\n\n\nREQUIREMENTS\n\n - Relevant experience: customer-embedded technical work, research engineering, technical founding experience, applied ML consulting with real build ownership, or end-to-end post-training programs.\n\n - You should be comfortable with senior external teams and able to connect experimental evidence to business decisions without lowering the research bar.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"8a447701-448a-4ddd-8607-bee8ac836d88","title":"Applied Research - Continual Learning","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:28:49.303+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/8a447701-448a-4ddd-8607-bee8ac836d88","applyUrl":"https://jobs.ashbyhq.com/primeintellect/8a447701-448a-4ddd-8607-bee8ac836d88/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will build continual learning systems that turn agent trajectories, corrections, and failures into new tasks and model updates. You will compare streaming RL, batched post-training, and distillation, and determine which improvements belong in model weights or persistent memory, skills, prompts, and subagent state. Your experiments will track skill acquisition, transfer, forgetting, and loss of plasticity over long sequences of tasks.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
Control update cadence, policy lag, task drift, and exploration, and evaluate each checkpoint on earlier skills, unseen tasks, and general capabilities. Build task-generation and training pipelines with reproducible data and harness versions, so regressions can be traced to a specific change and rolled back. You will work directly with the external collaborators whose deployments drive the research.
Representative early problems
Select high-value production traces and interventions, convert them into training signal, and establish regression gates before a change reaches deployment.
Determine which adaptation belongs in model weights, harness state, memory, tools, or data, then measure the tradeoffs across a sequence of tasks.
Build evaluations that distinguish retained learning and transfer from memorization, context accumulation, or measurement drift.
What you will own
Methods for selective consolidation, replay, adaptation, memory, and online or offline RL.
Evaluations for retention, transfer, regression, and rollback as tasks and distributions change.
Promotion gates and human oversight for agents that continue to change in production.
Field experiments, benchmarks, and open research that make continual improvement measurable and transferable.
Relevant experience: continual or online learning, RL, meta-learning, memory systems, post-training data loops, adaptive production ML, or evaluation under changing distributions.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou will build continual learning systems that turn agent trajectories, corrections, and failures into new tasks and model updates. You will compare streaming RL, batched post-training, and distillation, and determine which improvements belong in model weights or persistent memory, skills, prompts, and subagent state https://www.primeintellect.ai/blog/prime-agent. Your experiments will track skill acquisition, transfer, forgetting, and loss of plasticity over long sequences of tasks.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nControl update cadence, policy lag, task drift, and exploration, and evaluate each checkpoint on earlier skills, unseen tasks, and general capabilities. Build task-generation and training pipelines with reproducible data and harness versions, so regressions can be traced to a specific change and rolled back. You will work directly with the external collaborators whose deployments drive the research.\n\nRepresentative early problems\n\n - Select high-value production traces and interventions, convert them into training signal, and establish regression gates before a change reaches deployment.\n\n - Determine which adaptation belongs in model weights, harness state, memory, tools, or data, then measure the tradeoffs across a sequence of tasks.\n\n - Build evaluations that distinguish retained learning and transfer from memorization, context accumulation, or measurement drift.\n\nWhat you will own\n\n - Methods for selective consolidation, replay, adaptation, memory, and online or offline RL.\n\n - Evaluations for retention, transfer, regression, and rollback as tasks and distributions change.\n\n - Promotion gates and human oversight for agents that continue to change in production.\n\n - Field experiments, benchmarks, and open research that make continual improvement measurable and transferable.\n\n\nREQUIREMENTS\n\n - Relevant experience: continual or online learning, RL, meta-learning, memory systems, post-training data loops, adaptive production ML, or evaluation under changing distributions.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"8597d151-46db-4c67-a809-5d40daae7185","title":"Member of Technical Staff - Prime Agent","department":"Engineering","team":"Engineering","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-17T22:44:23.257+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/8597d151-46db-4c67-a809-5d40daae7185","applyUrl":"https://jobs.ashbyhq.com/primeintellect/8597d151-46db-4c67-a809-5d40daae7185/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You'll help build Prime Agent, our open-source coding and research agent. This is a generalist software engineering role spanning product development and the underlying agent systems. You'll own features end-to-end, from design through deployment and monitoring, and ship at the pace of an early-stage company tackling one of the most important problems in AI.
This role is based in our San Francisco office. Candidates must be based in San Francisco and work in person. This is not a remote role and is not open in other locations.
Build intuitive interfaces for working with agents and managing their tasks
Own features end-to-end, from design through deployment and operation
Work with users to understand their needs and turn feedback into product improvements
Work closely with teams across the Prime ecosystem to build cloud agents, self-improving loops, and integrations for traces and post-training
Work closely with researchers to turn new agent capabilities into reliable product features
Contribute to internal tooling, automation, and developer experience improvements
Strong software engineering skills in TypeScript and Python
Experience building developer tools, dashboards, platform products, and working with agent harnesses
Strong product judgment and attention to user experience, with the ability to turn complex capabilities into intuitive workflows
Experience building reliable systems with asynchronous execution and real-time interactions
Awareness of what's happening at the frontier - new models, agent capabilities, and tooling - and the ability to translate that into product decisions
A bias toward shipping, ownership of production code, and pragmatic engineering judgment
Experience building coding agents, agent harnesses, or developer tools powered by LLMs
CLI or terminal UI development
WebSocket / real-time systems experience
Open-source contributions
Prior experience at an early-stage startup
Cash compensation range of $150,000–$300,000 with significant equity incentives
In-person collaboration with the Prime Agent team in our San Francisco office
Full visa sponsorship and relocation support
Professional development budget for courses and conferences
Regular team off-sites and conference attendance
Opportunity to shape the future of open AI development
You'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source contributions.
We value potential over perfection - if you're a strong generalist engineer who is passionate about democratizing AI development, we want to talk to you.
Ready to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou'll help build Prime Agent, our open-source coding and research agent. This is a generalist software engineering role spanning product development and the underlying agent systems. You'll own features end-to-end, from design through deployment and monitoring, and ship at the pace of an early-stage company tackling one of the most important problems in AI.\n\nThis role is based in our San Francisco office. Candidates must be based in San Francisco and work in person. This is not a remote role and is not open in other locations.\n\n\nCORE RESPONSIBILITIES\n\n - Build intuitive interfaces for working with agents and managing their tasks\n\n - Own features end-to-end, from design through deployment and operation\n\n - Work with users to understand their needs and turn feedback into product improvements\n\n - Work closely with teams across the Prime ecosystem to build cloud agents, self-improving loops, and integrations for traces and post-training\n\n - Work closely with researchers to turn new agent capabilities into reliable product features\n\n - Contribute to internal tooling, automation, and developer experience improvements\n\n\nTECHNICAL REQUIREMENTS\n\n - Strong software engineering skills in TypeScript and Python\n\n - Experience building developer tools, dashboards, platform products, and working with agent harnesses\n\n - Strong product judgment and attention to user experience, with the ability to turn complex capabilities into intuitive workflows\n\n - Experience building reliable systems with asynchronous execution and real-time interactions\n\n - Awareness of what's happening at the frontier - new models, agent capabilities, and tooling - and the ability to translate that into product decisions\n\n - A bias toward shipping, ownership of production code, and pragmatic engineering judgment\n\n\nNICE TO HAVE\n\n - Experience building coding agents, agent harnesses, or developer tools powered by LLMs\n\n - CLI or terminal UI development\n\n - WebSocket / real-time systems experience\n\n - Open-source contributions\n\n - Prior experience at an early-stage startup\n\n\nWHAT WE OFFER\n\n - Cash compensation range of $150,000–$300,000 with significant equity incentives\n\n - In-person collaboration with the Prime Agent team in our San Francisco office\n\n - Full visa sponsorship and relocation support\n\n - Professional development budget for courses and conferences\n\n - Regular team off-sites and conference attendance\n\n - Opportunity to shape the future of open AI development\n\n\nGROWTH OPPORTUNITY\n\nYou'll join a team of experienced engineers and researchers working on cutting-edge problems in AI infrastructure. We believe in open development and encourage team members to contribute to the broader AI community through research and open-source contributions.\n\nWe value potential over perfection - if you're a strong generalist engineer who is passionate about democratizing AI development, we want to talk to you.\n\nReady to help shape the future of AI? Apply now and join us in our mission to make powerful AI models accessible to everyone."},{"id":"7bd4beca-6c73-4a38-9a6b-09dfa855dcf0","title":"Research Engineer / Scientist - Reinforcement Learning","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-10-16T13:54:28.845+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/7bd4beca-6c73-4a38-9a6b-09dfa855dcf0","applyUrl":"https://jobs.ashbyhq.com/primeintellect/7bd4beca-6c73-4a38-9a6b-09dfa855dcf0/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Help develop the algorithms and systems used to train Prime Intellect’s open models with reinforcement learning. We’re hiring research engineers and scientists across RL algorithms, scaling research, and training frameworks. Your focus will reflect your strengths; you don’t need to cover all three.
Algorithm research
Develop and test RL methods for reasoning, coding, and long-horizon tasks, including changes to training objectives and policy optimization.
Investigate stability in asynchronous RL, including policy staleness, trainer/sampler mismatch, and numerical differences between training and rollouts.
Work with data and evaluation researchers to understand how task selection, rewards, and training data affect what the model learns.
Scaling science
Design experiments to understand how RL performance and learning dynamics change with model size, training compute, and task horizon.
Use controlled ablations and evaluations to distinguish learning gains from reward exploitation, evaluation artifacts, or training bugs.
Turn experimental findings into training recipes, then test whether the improvements hold across larger models and longer runs.
RL frameworks
Extend prime-rl to support new algorithms, models, and asynchronous training workflows.
Improve the performance and numerical correctness of research runs across trainers, rollout workers, and environments, working with the infrastructure team.
Build diagnostics that make unstable runs and performance regressions easier to reproduce, and help distinguish algorithm problems from implementation or runtime issues.
Demonstrated depth in at least one of RL algorithms, empirical research on model training, or distributed ML systems. This could come through research, open-source contributions, or systems you’ve built.
Strong programming skills in Python and experience with PyTorch or a comparable ML framework.
Ability to design controlled experiments, interpret results, and debug unexpected model or system behavior.
Experience taking a research idea or systems change through implementation, testing, and evaluation, with a clear account of what worked and what didn’t.
Nice to have
Experience with distributed training frameworks or asynchronous rollout pipelines.
Experience diagnosing unstable RL runs, including policy-staleness or numerical issues.
Work on coding agents, verifiable rewards, or long-horizon environments.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n \n\n\nROLE IMPACT\n\nHelp develop the algorithms and systems used to train Prime Intellect’s open models with reinforcement learning. We’re hiring research engineers and scientists across RL algorithms, scaling research, and training frameworks. Your focus will reflect your strengths; you don’t need to cover all three.\n\n\nRESPONSIBILITIES\n\nAlgorithm research\n\n - Develop and test RL methods for reasoning, coding, and long-horizon tasks, including changes to training objectives and policy optimization.\n\n - Investigate stability in asynchronous RL, including policy staleness, trainer/sampler mismatch, and numerical differences between training and rollouts.\n\n - Work with data and evaluation researchers to understand how task selection, rewards, and training data affect what the model learns.\n\nScaling science\n\n - Design experiments to understand how RL performance and learning dynamics change with model size, training compute, and task horizon.\n\n - Use controlled ablations and evaluations to distinguish learning gains from reward exploitation, evaluation artifacts, or training bugs.\n\n - Turn experimental findings into training recipes, then test whether the improvements hold across larger models and longer runs.\n\nRL frameworks\n\n - Extend prime-rl https://github.com/PrimeIntellect-ai/prime-rl to support new algorithms, models, and asynchronous training workflows.\n\n - Improve the performance and numerical correctness of research runs across trainers, rollout workers, and environments, working with the infrastructure team.\n\n - Build diagnostics that make unstable runs and performance regressions easier to reproduce, and help distinguish algorithm problems from implementation or runtime issues.\n\n\nREQUIREMENTS\n\n - Demonstrated depth in at least one of RL algorithms, empirical research on model training, or distributed ML systems. This could come through research, open-source contributions, or systems you’ve built.\n\n - Strong programming skills in Python and experience with PyTorch or a comparable ML framework.\n\n - Ability to design controlled experiments, interpret results, and debug unexpected model or system behavior.\n\n - Experience taking a research idea or systems change through implementation, testing, and evaluation, with a clear account of what worked and what didn’t.\n\nNice to have\n\n - Experience with distributed training frameworks or asynchronous rollout pipelines.\n\n - Experience diagnosing unstable RL runs, including policy-staleness or numerical issues.\n\n - Work on coding agents, verifiable rewards, or long-horizon environments.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"eecf9af5-46fb-4b62-b8c8-921098fd6514","title":"Research Engineer - Pre-training Data","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-10-16T14:44:40.384+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/eecf9af5-46fb-4b62-b8c8-921098fd6514","applyUrl":"https://jobs.ashbyhq.com/primeintellect/eecf9af5-46fb-4b62-b8c8-921098fd6514/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Build the data behind Prime Intellect’s next generation of open models. You’ll find valuable sources, turn them into training-ready corpora, and work with researchers to test what actually improves the model. Your work will shape both broad pre-training and the data used to develop reasoning, coding, and long-context capabilities.
Find, evaluate, and collect valuable training data from web, code, and technical sources.
Build scalable pipelines that turn raw data into training-ready corpora through extraction, cleaning, deduplication, and filtering.
Develop checks for data quality, coverage, and contamination, and use them to improve collection and processing.
Work with researchers on data mixtures and sampling strategies for pre-training and midtraining.
Prepare and evaluate data for long-context learning, reasoning, and coding, testing synthetic data where it proves useful.
Run controlled data experiments and use model evaluations to decide which sources, mixtures, and processing choices are worth scaling.
Make large-scale data processing efficient, reproducible, and recoverable, with clear source provenance and dataset versioning.
Strong programming skills in Python and a record of building reliable data-processing or distributed systems.
Experience working with large datasets and managing the practical tradeoffs of storage, compute, throughput, and reproducibility.
Ability to diagnose data-quality issues and translate research goals into measurable checks and experiments.
Comfort owning an evolving technical problem end to end and working closely with researchers to test assumptions and improve the system.
Relevant experience may include:
Web crawling, document extraction, code-data collection, or large-scale corpus curation.
Deduplication, text classification, quality scoring, synthetic data, or training-data contamination analysis.
Data-mixture experiments, curriculum learning, or measuring how dataset changes affect model performance.
Nice to have
Experience preparing datasets for language-model pre-training, continued pre-training, or midtraining.
Work with distributed processing frameworks such as Spark, Ray, or comparable systems.
Experience with multilingual, mathematical, scientific, or repository-level code data.
Familiarity with tokenization, training data loaders, and the needs of large model-training runs.
Depth in data acquisition, scalable processing, or data research can be a strong starting point. We don’t expect every candidate to have worked across all of them.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nBuild the data behind Prime Intellect’s next generation of open models. You’ll find valuable sources, turn them into training-ready corpora, and work with researchers to test what actually improves the model. Your work will shape both broad pre-training and the data used to develop reasoning, coding, and long-context capabilities.\n\n\nRESPONSIBILITIES\n\n - Find, evaluate, and collect valuable training data from web, code, and technical sources.\n\n - Build scalable pipelines that turn raw data into training-ready corpora through extraction, cleaning, deduplication, and filtering.\n\n - Develop checks for data quality, coverage, and contamination, and use them to improve collection and processing.\n\n - Work with researchers on data mixtures and sampling strategies for pre-training and midtraining.\n\n - Prepare and evaluate data for long-context learning, reasoning, and coding, testing synthetic data where it proves useful.\n\n - Run controlled data experiments and use model evaluations to decide which sources, mixtures, and processing choices are worth scaling.\n\n - Make large-scale data processing efficient, reproducible, and recoverable, with clear source provenance and dataset versioning.\n\n\nREQUIREMENTS\n\n - Strong programming skills in Python and a record of building reliable data-processing or distributed systems.\n\n - Experience working with large datasets and managing the practical tradeoffs of storage, compute, throughput, and reproducibility.\n\n - Ability to diagnose data-quality issues and translate research goals into measurable checks and experiments.\n\n - Comfort owning an evolving technical problem end to end and working closely with researchers to test assumptions and improve the system.\n\nRelevant experience may include:\n\n - Web crawling, document extraction, code-data collection, or large-scale corpus curation.\n\n - Deduplication, text classification, quality scoring, synthetic data, or training-data contamination analysis.\n\n - Data-mixture experiments, curriculum learning, or measuring how dataset changes affect model performance.\n\nNice to have\n\n - Experience preparing datasets for language-model pre-training, continued pre-training, or midtraining.\n\n - Work with distributed processing frameworks such as Spark, Ray, or comparable systems.\n\n - Experience with multilingual, mathematical, scientific, or repository-level code data.\n\n - Familiarity with tokenization, training data loaders, and the needs of large model-training runs.\n\nDepth in data acquisition, scalable processing, or data research can be a strong starting point. We don’t expect every candidate to have worked across all of them.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"7ef14297-51c3-45b1-abe4-890489a0bddb","title":"Research Engineer - Data and Evaluation","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-10-16T14:44:39.894+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/7ef14297-51c3-45b1-abe4-890489a0bddb","applyUrl":"https://jobs.ashbyhq.com/primeintellect/7ef14297-51c3-45b1-abe4-890489a0bddb/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Build the tasks and environments that train Prime Intellect’s models through RL, and the evaluations that show what they can actually do. You’ll work directly with researchers to investigate model failures, from how a task is implemented to how success is scored.
Design and build challenging evaluations for coding and long-horizon agents.
Build executable environments in Verifiers, with realistic task setup and clear success criteria.
Develop and test scoring logic and verifiers, investigating reward hacks and misleading measures of progress.
Trace failures through model outputs and harness code to distinguish model limitations from problems in the environment or scoring.
Build and scale RL data pipelines that generate and curate challenging coding and agentic tasks.
Work with RL researchers to test whether changes to tasks and training data improve performance on held-out tasks.
Strong software engineering skills in Python and experience building data pipelines, evaluation systems, or ML workflows used by others.
Hands-on experience working with language models or agents, with an understanding of how training data and evaluation choices affect model behavior.
Ability to inspect complex data and model outputs, identify failure modes, and turn an ambiguous quality question into a testable measurement.
A record of owning technical projects through experimentation, implementation, and validation, while collaborating closely with researchers and engineers.
Hands-on experience building technical evaluations or agent environments, with the engineering depth to inspect model traces and debug harness and scoring code.
Relevant experience may include:
Large-scale synthetic data generation, dataset curation, or distributed processing for model training.
Evaluation methodology, executable coding tasks, agent environments, verifiers, or reward design.
Experiment analysis, statistical testing, contamination detection, and reproducible research tooling.
Nice to have
Experience with RL or post-training data, including task difficulty selection and analysis of rollout or reward distributions.
Contributions to Verifiers, Prime-RL, or comparable environment and evaluation frameworks.
Work on model-based grading, long-horizon coding agents, or diagnosing reward exploitation.
We welcome depth in data generation, environments and rewards, or evaluation systems. You don’t need to have worked across every area.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nBuild the tasks and environments that train Prime Intellect’s models through RL, and the evaluations that show what they can actually do. You’ll work directly with researchers to investigate model failures, from how a task is implemented to how success is scored.\n\n\nRESPONSIBILITIES\n\n - Design and build challenging evaluations for coding and long-horizon agents.\n\n - Build executable environments in Verifiers, with realistic task setup and clear success criteria.\n\n - Develop and test scoring logic and verifiers, investigating reward hacks and misleading measures of progress.\n\n - Trace failures through model outputs and harness code to distinguish model limitations from problems in the environment or scoring.\n\n - Build and scale RL data pipelines that generate and curate challenging coding and agentic tasks.\n\n - Work with RL researchers to test whether changes to tasks and training data improve performance on held-out tasks.\n\n\nREQUIREMENTS\n\n - Strong software engineering skills in Python and experience building data pipelines, evaluation systems, or ML workflows used by others.\n\n - Hands-on experience working with language models or agents, with an understanding of how training data and evaluation choices affect model behavior.\n\n - Ability to inspect complex data and model outputs, identify failure modes, and turn an ambiguous quality question into a testable measurement.\n\n - A record of owning technical projects through experimentation, implementation, and validation, while collaborating closely with researchers and engineers.\n\n - Hands-on experience building technical evaluations or agent environments, with the engineering depth to inspect model traces and debug harness and scoring code.\n\nRelevant experience may include:\n\n - Large-scale synthetic data generation, dataset curation, or distributed processing for model training.\n\n - Evaluation methodology, executable coding tasks, agent environments, verifiers, or reward design.\n\n - Experiment analysis, statistical testing, contamination detection, and reproducible research tooling.\n\nNice to have\n\n - Experience with RL or post-training data, including task difficulty selection and analysis of rollout or reward distributions.\n\n - Contributions to Verifiers, Prime-RL, or comparable environment and evaluation frameworks.\n\n - Work on model-based grading, long-horizon coding agents, or diagnosing reward exploitation.\n\nWe welcome depth in data generation, environments and rewards, or evaluation systems. You don’t need to have worked across every area.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"3103e1a4-2515-4ae0-a2f0-784e25d30cfe","title":"Research Engineer - Inference","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-10-16T14:44:28.971+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/3103e1a4-2515-4ae0-a2f0-784e25d30cfe","applyUrl":"https://jobs.ashbyhq.com/primeintellect/3103e1a4-2515-4ae0-a2f0-784e25d30cfe/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Build the inference systems that power Prime Intellect’s RL research, scaling toward 1–3T parameter models on NVL72 clusters. Your work will improve rollout throughput and make frequent model updates practical.
Extend prime-rl’s vLLM integration to bring new model architectures into RL training, with small-scale validation before larger runs.
Improve rollout throughput for long-horizon agents, tuning concurrency and KV-cache reuse as context lengths grow.
Optimize multi-node MoE inference, including expert parallelism and the tradeoffs of separating prefill from decode.
Reduce trainer-to-inference weight-update overhead through sharding-aware transfer and communication overlap, while keeping rollout workers on the intended policy version.
Build on FP8 and evaluate lower precision. Use trainer/sampler mismatch checks to catch numerical regressions and measure the effect on RL training.
Work with model researchers on architecture choices that improve inference efficiency, balancing model quality against execution cost on the target hardware.
Experience building or optimizing GPU inference systems, with an understanding of inference-engine internals.
Strong Python skills and experience with PyTorch or a comparable ML framework.
Ability to profile end-to-end workloads and turn measurements into verified performance improvements.
Experience debugging numerical or integration issues in ML systems.
Nice to have
Contributions to vLLM or another inference engine.
CUDA or Triton kernel work, particularly for low-precision inference.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nBuild the inference systems that power Prime Intellect’s RL research, scaling toward 1–3T parameter models on NVL72 clusters. Your work will improve rollout throughput and make frequent model updates practical.\n\n\nRESPONSIBILITIES\n\n - Extend prime-rl’s vLLM integration to bring new model architectures into RL training, with small-scale validation before larger runs.\n\n - Improve rollout throughput for long-horizon agents, tuning concurrency and KV-cache reuse as context lengths grow.\n\n - Optimize multi-node MoE inference, including expert parallelism and the tradeoffs of separating prefill from decode.\n\n - Reduce trainer-to-inference weight-update overhead through sharding-aware transfer and communication overlap, while keeping rollout workers on the intended policy version.\n\n - Build on FP8 and evaluate lower precision. Use trainer/sampler mismatch checks to catch numerical regressions and measure the effect on RL training.\n\n - Work with model researchers on architecture choices that improve inference efficiency, balancing model quality against execution cost on the target hardware.\n\n\nREQUIREMENTS\n\n - Experience building or optimizing GPU inference systems, with an understanding of inference-engine internals.\n\n - Strong Python skills and experience with PyTorch or a comparable ML framework.\n\n - Ability to profile end-to-end workloads and turn measurements into verified performance improvements.\n\n - Experience debugging numerical or integration issues in ML systems.\n\nNice to have\n\n - Contributions to vLLM or another inference engine.\n\n - CUDA or Triton kernel work, particularly for low-precision inference.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"9dae2470-0e2c-4d42-8e7f-20b7f534440e","title":"Research Engineer - Pre-training","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}}],"publishedAt":"2026-10-16T14:44:39.749+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/9dae2470-0e2c-4d42-8e7f-20b7f534440e","applyUrl":"https://jobs.ashbyhq.com/primeintellect/9dae2470-0e2c-4d42-8e7f-20b7f534440e/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Take Prime Intellect’s next open models from architecture experiments through full pre-training runs, scaling toward 1–3T parameters. You’ll work across the model and training stack, making research decisions and fixing the systems problems that determine whether a run succeeds.
Build and extend the pre-training code in prime-rl, taking small experiments into full training runs.
Evaluate model architecture changes, including MoE and attention designs, against model quality and training cost.
Profile training jobs and improve end-to-end throughput, working with systems engineers on hardware-aware execution.
Make long runs resilient to failures, with reliable checkpointing and recovery.
Investigate training dynamics and numerical stability, checking that optimizations preserve convergence.
Run controlled ablations to guide larger experiments, keeping the setup and results reproducible.
Work with data and evaluation researchers on training mixtures, including midtraining and long-context experiments.
Hands-on experience training language models, including modifying training code and following the effects through actual runs.
Strong Python skills and experience with PyTorch or a comparable framework, with enough systems understanding to debug distributed training.
A practical understanding of model architectures and optimization, including their effect on convergence.
Ability to design an ablation that answers a research question and explain what its results do and don’t show.
Good judgment about what to investigate yourself and when to bring in a colleague.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators, including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nTake Prime Intellect’s next open models from architecture experiments through full pre-training runs, scaling toward 1–3T parameters. You’ll work across the model and training stack, making research decisions and fixing the systems problems that determine whether a run succeeds.\n\n\nRESPONSIBILITIES\n\n - Build and extend the pre-training code in prime-rl, taking small experiments into full training runs.\n\n - Evaluate model architecture changes, including MoE and attention designs, against model quality and training cost.\n\n - Profile training jobs and improve end-to-end throughput, working with systems engineers on hardware-aware execution.\n\n - Make long runs resilient to failures, with reliable checkpointing and recovery.\n\n - Investigate training dynamics and numerical stability, checking that optimizations preserve convergence.\n\n - Run controlled ablations to guide larger experiments, keeping the setup and results reproducible.\n\n - Work with data and evaluation researchers on training mixtures, including midtraining and long-context experiments.\n\n\nREQUIREMENTS\n\n - Hands-on experience training language models, including modifying training code and following the effects through actual runs.\n\n - Strong Python skills and experience with PyTorch or a comparable framework, with enough systems understanding to debug distributed training.\n\n - A practical understanding of model architectures and optimization, including their effect on convergence.\n\n - Ability to design an ablation that answers a research question and explain what its results do and don’t show.\n\n - Good judgment about what to investigate yourself and when to bring in a colleague.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"7f41cc90-9bf1-49a8-9285-3838033e627d","title":"Applied Research - Scientific AI","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"Berlin","address":{"postalAddress":{"addressRegion":"Berlin","addressCountry":"Germany","addressLocality":"Berlin"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T16:30:05.734+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/7f41cc90-9bf1-49a8-9285-3838033e627d","applyUrl":"https://jobs.ashbyhq.com/primeintellect/7f41cc90-9bf1-49a8-9285-3838033e627d/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will build and post-train agents for scientific research: searching literature, writing and running code, using domain tools, and choosing the next experiment. The work spans autonomous research and computational science, with environments and evaluations built around research partners' tools, data, and experimental feedback.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Prime's autonomous-research study used multi-day nanoGPT optimization tasks to examine how agents measure noise, revisit earlier experiments, and verify improvements. You will build similar experimental loops for scientific tools and datasets.
Build long-horizon environments where agents search literature, query scientific databases, run code and domain tools, and choose experiments within a fixed compute or experimental budget.
Turn research partners' workflows, datasets, and outcome criteria into trainable tasks, evaluations, and reward signals.
Design verifiers for correctness, novelty, and reproducibility. Test whether agents use the underlying evidence, with checks for label leakage, lookup shortcuts, and unsupported conclusions.
Connect programmatic checks, simulations, domain models, and expert feedback to training. Account for noisy or delayed results, calibrate confidence estimates, and define where an experiment needs human review.
Develop RL and multi-agent methods for proposing, testing, and critiquing hypotheses. Compare them with strong single-agent baselines at equal compute and evaluate on held-out problems.
Keep code, data, logs, and experiment configurations reproducible. Re-run promising results, measure noise, and use ablations to separate model improvements from changes to the data, harness, or inference setup. Release reusable environments, evaluations, and research code.
We are looking first for depth in RL, agents, or post-training systems and the ability to acquire the relevant scientific context.
Prior work in autonomous research, ML for science, scientific computing, or reproducible computational research is useful but not required.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, Berlin, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou will build and post-train agents for scientific research: searching literature, writing and running code, using domain tools, and choosing the next experiment. The work spans autonomous research and computational science, with environments and evaluations built around research partners' tools, data, and experimental feedback.\n\n\nRESPONSIBILITIES\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nPrime's autonomous-research study https://www.primeintellect.ai/blog/measuring-autonomous-research used multi-day nanoGPT optimization tasks to examine how agents measure noise, revisit earlier experiments, and verify improvements. You will build similar experimental loops for scientific tools and datasets.\n\n - Build long-horizon environments where agents search literature, query scientific databases, run code and domain tools, and choose experiments within a fixed compute or experimental budget.\n\n - Turn research partners' workflows, datasets, and outcome criteria into trainable tasks, evaluations, and reward signals.\n\n - Design verifiers for correctness, novelty, and reproducibility. Test whether agents use the underlying evidence, with checks for label leakage, lookup shortcuts, and unsupported conclusions.\n\n - Connect programmatic checks, simulations, domain models, and expert feedback to training. Account for noisy or delayed results, calibrate confidence estimates, and define where an experiment needs human review.\n\n - Develop RL and multi-agent methods for proposing, testing, and critiquing hypotheses. Compare them with strong single-agent baselines at equal compute and evaluate on held-out problems.\n\n - Keep code, data, logs, and experiment configurations reproducible. Re-run promising results, measure noise, and use ablations to separate model improvements from changes to the data, harness, or inference setup. Release reusable environments, evaluations, and research code.\n\n\nREQUIREMENTS\n\n - We are looking first for depth in RL, agents, or post-training systems and the ability to acquire the relevant scientific context.\n\n - Prior work in autonomous research, ML for science, scientific computing, or reproducible computational research is useful but not required.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, Berlin, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"1aec9bbb-74d2-4353-9be3-207f15ec6a4b","title":"Applied Research - Sovereign AI","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:30:25.316+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/1aec9bbb-74d2-4353-9be3-207f15ec6a4b","applyUrl":"https://jobs.ashbyhq.com/primeintellect/1aec9bbb-74d2-4353-9be3-207f15ec6a4b/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
National institutions, public-sector organizations, and regulated industries need capable models and agents that operate within strict boundaries around data, deployment, access, auditability, and human authority. They also need the ability to improve those systems on their own workflows rather than depend indefinitely on a closed external model. You will build post-training and agent systems in public-sector and regulated-industry settings.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
The work happens inside partner constraints, sometimes on their infrastructure and under their access rules.
Representative early problems
Build a private improvement loop that turns sensitive traces into environments, training signal, evaluations, and deployable model updates without crossing the operator's data boundary.
Train a tool-using domain agent whose actions are permissioned, traceable, recoverable, and evaluated against real procedures rather than generic benchmarks.
Develop high-assurance evaluations for reliability, calibration, robustness, human escalation, and compliance with domain constraints under adversarial or changing conditions.
What you will own
Improvement loops that run end to end inside the operator's boundary, from trace selection to deployed update.
Agents with explicit permissions, evidence trails, bounded tools, and human checkpoints.
Evaluations that test reliability, calibration, recovery, and compliance with real procedures.
Experiments on how privacy, access, and infrastructure constraints change learning, with methods that transfer across high-assurance settings.
Relevant experience: RL/post-training, private ML systems, high-assurance evaluation, agent safety, regulated-domain ML, public-sector technical work, or deployment under strict data and infrastructure constraints.
You should want to build and evaluate these systems directly.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nNational institutions, public-sector organizations, and regulated industries need capable models and agents that operate within strict boundaries around data, deployment, access, auditability, and human authority. They also need the ability to improve those systems on their own workflows rather than depend indefinitely on a closed external model. You will build post-training and agent systems in public-sector and regulated-industry settings.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nThe work happens inside partner constraints, sometimes on their infrastructure and under their access rules.\n\nRepresentative early problems\n\n - Build a private improvement loop that turns sensitive traces into environments, training signal, evaluations, and deployable model updates without crossing the operator's data boundary.\n\n - Train a tool-using domain agent whose actions are permissioned, traceable, recoverable, and evaluated against real procedures rather than generic benchmarks.\n\n - Develop high-assurance evaluations for reliability, calibration, robustness, human escalation, and compliance with domain constraints under adversarial or changing conditions.\n\nWhat you will own\n\n - Improvement loops that run end to end inside the operator's boundary, from trace selection to deployed update.\n\n - Agents with explicit permissions, evidence trails, bounded tools, and human checkpoints.\n\n - Evaluations that test reliability, calibration, recovery, and compliance with real procedures.\n\n - Experiments on how privacy, access, and infrastructure constraints change learning, with methods that transfer across high-assurance settings.\n\n\nREQUIREMENTS\n\n - Relevant experience: RL/post-training, private ML systems, high-assurance evaluation, agent safety, regulated-domain ML, public-sector technical work, or deployment under strict data and infrastructure constraints.\n\n - You should want to build and evaluate these systems directly.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"9ed2340b-029e-4f58-82f5-0d55f09aad5b","title":"Applied Research - Chip Design","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:28:27.247+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/9ed2340b-029e-4f58-82f5-0d55f09aad5b","applyUrl":"https://jobs.ashbyhq.com/primeintellect/9ed2340b-029e-4f58-82f5-0d55f09aad5b/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Chip design is a natural fit for agentic reinforcement learning: long-horizon work, specialized tools, expensive mistakes, hard constraints, and objective feedback from simulation, verification, timing, power, and area. You will build agents, environments, evaluations, and post-training systems for real semiconductor design and verification workflows. The work is grounded in an external design workflow, with access to relevant domain experts and toolchains. It may span RTL generation and debugging, formal verification, simulation, physical design, optimization, and orchestration across EDA tools.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Representative early problems
Build an interactive verification environment where an agent reads a specification, writes or repairs RTL, invokes simulation or formal tools, interprets failures, and iterates toward a verified result.
Design rewards and evaluations that combine functional correctness with tool efficiency, robustness, and performance/power/area constraints without opening easy shortcuts.
Train long-horizon agents to recover from design and tool failures, preserve useful state, and generalize across blocks, designs, and constraint sets.
What you will own
Interactive environments where agents use EDA tools to modify designs, debug failures, and plan multi-step work.
Verifiers and rewards grounded in tests, simulation, formal methods, design rules, and engineering outcomes.
Experiments on credit assignment when feedback arrives late and several tools contribute.
Evaluations for correctness, recovery, and generalization, with reusable methods for other simulator-grounded engineering domains.
We are looking first for depth in RL, agents, or post-training systems and the ability to become fluent in semiconductor workflows.
Prior work in design automation, compilers, formal methods, or ML for systems is useful but not required.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nChip design is a natural fit for agentic reinforcement learning: long-horizon work, specialized tools, expensive mistakes, hard constraints, and objective feedback from simulation, verification, timing, power, and area. You will build agents, environments, evaluations, and post-training systems for real semiconductor design and verification workflows. The work is grounded in an external design workflow, with access to relevant domain experts and toolchains. It may span RTL generation and debugging, formal verification, simulation, physical design, optimization, and orchestration across EDA tools.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRepresentative early problems\n\n - Build an interactive verification environment where an agent reads a specification, writes or repairs RTL, invokes simulation or formal tools, interprets failures, and iterates toward a verified result.\n\n - Design rewards and evaluations that combine functional correctness with tool efficiency, robustness, and performance/power/area constraints without opening easy shortcuts.\n\n - Train long-horizon agents to recover from design and tool failures, preserve useful state, and generalize across blocks, designs, and constraint sets.\n\nWhat you will own\n\n - Interactive environments where agents use EDA tools to modify designs, debug failures, and plan multi-step work.\n\n - Verifiers and rewards grounded in tests, simulation, formal methods, design rules, and engineering outcomes.\n\n - Experiments on credit assignment when feedback arrives late and several tools contribute.\n\n - Evaluations for correctness, recovery, and generalization, with reusable methods for other simulator-grounded engineering domains.\n\n\nREQUIREMENTS\n\n - We are looking first for depth in RL, agents, or post-training systems and the ability to become fluent in semiconductor workflows.\n\n - Prior work in design automation, compilers, formal methods, or ML for systems is useful but not required.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"93783399-a5cc-4037-853e-824cf0ad1aeb","title":"Applied Research - Robotics","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:30:19.614+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/93783399-a5cc-4037-853e-824cf0ad1aeb","applyUrl":"https://jobs.ashbyhq.com/primeintellect/93783399-a5cc-4037-853e-824cf0ad1aeb/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
You will build environments, evaluations, agent systems, and post-training methods with an external robotics team. You will train and evaluate agents with reinforcement learning, world modeling, imitation learning, planning, and multimodal perception, using feedback from simulation and deployed robots. The experiments will test behavior under partial observations, long horizons, changing environments, and hardware and safety constraints.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Representative early problems
Build aligned simulation and real-world evaluations for a long-horizon embodied task, then measure exactly where policies fail to transfer.
Combine demonstrations, interventions, offline data, and online feedback into a training loop that improves behavior without erasing prior competence.
Train an embodied agent to recover from partial failures, changing scenes, and uncertain state while respecting safety and hardware constraints.
What you will own
Simulation and real-world environments for training, evaluation, and careful sim-to-real comparison.
Long-horizon agents that perceive, plan, act, remember, and recover under hardware constraints.
Training that combines offline and online RL, demonstrations, interventions, and outcome feedback.
Rewards and evaluations for task success, safety, adaptation, and continual improvement without silent regressions.
Relevant experience: robotics, embodied AI, RL, imitation learning, control, world models, simulation, computer vision, agent evaluation, or production systems interacting with the physical world.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nYou will build environments, evaluations, agent systems, and post-training methods with an external robotics team. You will train and evaluate agents with reinforcement learning, world modeling, imitation learning, planning, and multimodal perception, using feedback from simulation and deployed robots. The experiments will test behavior under partial observations, long horizons, changing environments, and hardware and safety constraints.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRepresentative early problems\n\n - Build aligned simulation and real-world evaluations for a long-horizon embodied task, then measure exactly where policies fail to transfer.\n\n - Combine demonstrations, interventions, offline data, and online feedback into a training loop that improves behavior without erasing prior competence.\n\n - Train an embodied agent to recover from partial failures, changing scenes, and uncertain state while respecting safety and hardware constraints.\n\nWhat you will own\n\n - Simulation and real-world environments for training, evaluation, and careful sim-to-real comparison.\n\n - Long-horizon agents that perceive, plan, act, remember, and recover under hardware constraints.\n\n - Training that combines offline and online RL, demonstrations, interventions, and outcome feedback.\n\n - Rewards and evaluations for task success, safety, adaptation, and continual improvement without silent regressions.\n\n\nREQUIREMENTS\n\n - Relevant experience: robotics, embodied AI, RL, imitation learning, control, world models, simulation, computer vision, agent evaluation, or production systems interacting with the physical world.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"e6430d82-f266-475e-a273-255c5b2bd521","title":"Applied Research - Quantitative Finance","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:29:38.800+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/e6430d82-f266-475e-a273-255c5b2bd521","applyUrl":"https://jobs.ashbyhq.com/primeintellect/e6430d82-f266-475e-a273-255c5b2bd521/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Quantitative finance is a demanding setting for agentic reinforcement learning: observations are noisy and non-stationary, actions have delayed and path-dependent consequences, naive objectives invite exploitation, and a convincing historical result can disappear under costs, constraints, or a new regime. You will build environments, evaluations, agents, and post-training methods for quantitative research and systematic decision-making. We have active programs with quantitative funds and other finance teams. The work may span research agents, market and portfolio simulators, time-series reasoning, execution, risk, and the production feedback loops that connect hypotheses to deployed systems.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Representative early problems
Build a quantitative-research environment where an agent forms a hypothesis, writes and runs analysis, runs a backtest on point-in-time data, diagnoses the result, and decides what to test next.
Design rewards and evaluations that distinguish durable improvement from the artifacts of look-ahead bias, data leakage, overfitting, favorable execution assumptions, or hidden concentration and tail risk.
Train long-horizon agents to use data, research, coding, simulation, and portfolio tools while preserving provenance, respecting risk constraints, and adapting when market conditions change.
What you will own
Environments grounded in point-in-time data, realistic market mechanics, transaction costs, latency, capacity, and operational constraints.
Agents that conduct quantitative research, use analytical and execution tools, inspect their evidence, and recover from failed hypotheses.
Verifiers and evaluations that test robust out-of-sample behavior across assets, horizons, regimes, and adverse conditions.
Production loops with monitoring, human authority, regression gates, and failure attribution across models, data, simulation, and infrastructure, producing methods that transfer to other noisy, high-stakes decision systems.
We are looking first for depth in RL, agents, or post-training systems and the ability to become fluent in quantitative research.
Prior work in systematic trading, market microstructure, time-series modeling, econometrics, or financial simulation is useful but not required.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nQuantitative finance is a demanding setting for agentic reinforcement learning: observations are noisy and non-stationary, actions have delayed and path-dependent consequences, naive objectives invite exploitation, and a convincing historical result can disappear under costs, constraints, or a new regime. You will build environments, evaluations, agents, and post-training methods for quantitative research and systematic decision-making. We have active programs with quantitative funds and other finance teams. The work may span research agents, market and portfolio simulators, time-series reasoning, execution, risk, and the production feedback loops that connect hypotheses to deployed systems.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRepresentative early problems\n\n - Build a quantitative-research environment where an agent forms a hypothesis, writes and runs analysis, runs a backtest on point-in-time data, diagnoses the result, and decides what to test next.\n\n - Design rewards and evaluations that distinguish durable improvement from the artifacts of look-ahead bias, data leakage, overfitting, favorable execution assumptions, or hidden concentration and tail risk.\n\n - Train long-horizon agents to use data, research, coding, simulation, and portfolio tools while preserving provenance, respecting risk constraints, and adapting when market conditions change.\n\nWhat you will own\n\n - Environments grounded in point-in-time data, realistic market mechanics, transaction costs, latency, capacity, and operational constraints.\n\n - Agents that conduct quantitative research, use analytical and execution tools, inspect their evidence, and recover from failed hypotheses.\n\n - Verifiers and evaluations that test robust out-of-sample behavior across assets, horizons, regimes, and adverse conditions.\n\n - Production loops with monitoring, human authority, regression gates, and failure attribution across models, data, simulation, and infrastructure, producing methods that transfer to other noisy, high-stakes decision systems.\n\n\nREQUIREMENTS\n\n - We are looking first for depth in RL, agents, or post-training systems and the ability to become fluent in quantitative research.\n\n - Prior work in systematic trading, market microstructure, time-series modeling, econometrics, or financial simulation is useful but not required.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"a6448ceb-aa39-4661-a2c0-350ef365c913","title":"Applied Research - Computer Use","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:28:42.917+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/a6448ceb-aa39-4661-a2c0-350ef365c913","applyUrl":"https://jobs.ashbyhq.com/primeintellect/a6448ceb-aa39-4661-a2c0-350ef365c913/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Computer-use agents must understand changing interfaces, maintain intent over long workflows, recover from mistakes, manage permissions and side effects, and know when an outcome is actually complete. You will build post-training and evaluation systems for multimodal agents operating browsers, desktops, and cross-application workflows. The work is grounded in an active external program. The research challenge is turning real workflows and traces into environments, rewards, and experiments that produce reliable improvement.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Representative early problems
Build long-horizon computer-use environments from real workflows while preventing shortcuts, leakage, and unsafe side effects.
Train agents to recover from interface changes, partial failures, and ambiguous intermediate states instead of restarting or silently drifting.
Develop evaluations that distinguish genuine completion from plausible-looking behavior, including correctness, efficiency, permissions, and appropriate escalation.
What you will own
Multimodal environments built from partner workflows, demonstrations, tools, interface state, and outcome criteria.
Browser, desktop, and cross-application agents with reliable sandbox and boundary enforcement.
Training and evaluation from demonstrations, interventions, production failures, synthetic tasks, and verifiable outcomes.
Experiments on recovery and generalization across applications, layouts, permissions, and deployment contexts.
Relevant experience: multimodal agents, computer use, browser or desktop automation, RL, imitation learning, HCI, embodied agents, or long-horizon agent evaluation.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nComputer-use agents must understand changing interfaces, maintain intent over long workflows, recover from mistakes, manage permissions and side effects, and know when an outcome is actually complete. You will build post-training and evaluation systems for multimodal agents operating browsers, desktops, and cross-application workflows. The work is grounded in an active external program. The research challenge is turning real workflows and traces into environments, rewards, and experiments that produce reliable improvement.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRepresentative early problems\n\n - Build long-horizon computer-use environments from real workflows while preventing shortcuts, leakage, and unsafe side effects.\n\n - Train agents to recover from interface changes, partial failures, and ambiguous intermediate states instead of restarting or silently drifting.\n\n - Develop evaluations that distinguish genuine completion from plausible-looking behavior, including correctness, efficiency, permissions, and appropriate escalation.\n\nWhat you will own\n\n - Multimodal environments built from partner workflows, demonstrations, tools, interface state, and outcome criteria.\n\n - Browser, desktop, and cross-application agents with reliable sandbox and boundary enforcement.\n\n - Training and evaluation from demonstrations, interventions, production failures, synthetic tasks, and verifiable outcomes.\n\n - Experiments on recovery and generalization across applications, layouts, permissions, and deployment contexts.\n\n\nREQUIREMENTS\n\n - Relevant experience: multimodal agents, computer use, browser or desktop automation, RL, imitation learning, HCI, embodied agents, or long-horizon agent evaluation.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"d45f07a8-a506-4e54-8640-23ad76cc9a2e","title":"Applied Research - Ecosystem & Product","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[{"location":"New York","address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}}},{"location":"Remote","address":{"postalAddress":{"addressCountry":"United States"}}},{"location":"London","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}}}],"publishedAt":"2026-10-19T15:29:07.133+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/d45f07a8-a506-4e54-8640-23ad76cc9a2e","applyUrl":"https://jobs.ashbyhq.com/primeintellect/d45f07a8-a506-4e54-8640-23ad76cc9a2e/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
Prime's research compounds when other researchers can run it, inspect it, extend it, and build on it. You will own the path from emerging research artifact to excellent open-source and product experience. This role sits inside Applied Research and works deeply with engineering. You will build core open-source systems, shape researcher-facing features in our training platform, decide which abstractions deserve stable interfaces, and help external contributors ship serious work into the stack.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
You will work continuously with external researchers and contributors. Prime's public surface already includes verifiers v1, prime-rl's extensible algorithms layer, and more than 2,500 community environments. The opportunity is to make the full path from task and harness through evaluation, training, and deployment feel like one coherent research system.
Representative early problems
Make a new researcher successful from environment idea to credible evaluation and training run without hiding the underlying experimental choices.
Turn fast-moving internal agent and RL abstractions into stable public interfaces without freezing weak ideas prematurely.
Create a contributor path where external researchers can ship environments, harnesses, algorithms, and integrations that meet a serious technical bar.
What you will own
Important open-source RL, environment, evaluation, and agent systems.
Researcher-facing features in our training platform, built with engineering and design.
APIs, examples, reference implementations, and technical explanations that expose useful abstractions.
Work with research users and external contributors that leads to better interfaces, inspectable code, and research artifacts.
Relevant experience: ownership of open-source ML or developer-tool systems, research engineering, technical product work, or communities formed around code and reproducible research.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Flexible Work (San Francisco, New York, London, or remote)
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nPrime's research compounds when other researchers can run it, inspect it, extend it, and build on it. You will own the path from emerging research artifact to excellent open-source and product experience. This role sits inside Applied Research and works deeply with engineering. You will build core open-source systems, shape researcher-facing features in our training platform, decide which abstractions deserve stable interfaces, and help external contributors ship serious work into the stack.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nYou will work continuously with external researchers and contributors. Prime's public surface already includes verifiers v1 https://www.primeintellect.ai/blog/verifiers-v1, prime-rl's extensible algorithms layer https://www.primeintellect.ai/blog/algorithms-layer, and more than 2,500 community environments. The opportunity is to make the full path from task and harness through evaluation, training, and deployment feel like one coherent research system.\n\nRepresentative early problems\n\n - Make a new researcher successful from environment idea to credible evaluation and training run without hiding the underlying experimental choices.\n\n - Turn fast-moving internal agent and RL abstractions into stable public interfaces without freezing weak ideas prematurely.\n\n - Create a contributor path where external researchers can ship environments, harnesses, algorithms, and integrations that meet a serious technical bar.\n\nWhat you will own\n\n - Important open-source RL, environment, evaluation, and agent systems.\n\n - Researcher-facing features in our training platform, built with engineering and design.\n\n - APIs, examples, reference implementations, and technical explanations that expose useful abstractions.\n\n - Work with research users and external contributors that leads to better interfaces, inspectable code, and research artifacts.\n\n\nREQUIREMENTS\n\n - Relevant experience: ownership of open-source ML or developer-tool systems, research engineering, technical product work, or communities formed around code and reproducible research.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Flexible Work (San Francisco, New York, London, or remote)\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"d63b3277-651e-4a81-85d8-888f0aef14f5","title":"Applied Research Lead - New York","department":"Applied Research","team":"Applied Research","employmentType":"FullTime","location":"New York","secondaryLocations":[],"publishedAt":"2026-10-19T15:29:31.942+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"New York","addressCountry":"United States","addressLocality":"New York City"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/d63b3277-651e-4a81-85d8-888f0aef14f5","applyUrl":"https://jobs.ashbyhq.com/primeintellect/d63b3277-651e-4a81-85d8-888f0aef14f5/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
We are looking for a senior regional player-coach who can establish that group through personally led research and field programs. You will begin close to the work: shape consequential external problems into coherent research programs, lead the hardest early experiments and deployments, and recruit exceptional researchers who want the same combination of frontier methods and real-world responsibility. As the group grows, you will set its research direction and the technical and experimental standards that let it move quickly without losing rigor.
Applied Research at Prime
We turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.
Every Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.
Role scope
This is a New York leadership role within Prime's broader Applied Research function. It requires research taste, technical depth, customer and partner judgment, exceptional recruiting, and the willingness to remain a builder. You will stay directly involved with external program teams while leading technical work, recruiting, and developing the New York group.
Representative early problems
Turn the first portfolio of New York field programs into research agendas with strong owners, comparable evidence, and explicit paths from domain result to shared capability.
Personally lead a high-stakes program through environment design, baseline, post-training, evaluation, and deployment while establishing the quality bar for the growing team.
Recruit the founding New York researchers and establish a working rhythm that keeps review rigorous without slowing the team.
What you will own
A small number of the hardest New York partner and customer research programs.
Recruiting and development of the founding New York Applied Research group.
Technical standards for environments, evaluations, post-training experiments, failure analysis, and deployed results.
Meaningful personal research, cross-program technical direction, and decisions about what becomes shared research or product capability as the team grows.
A track record leading high-performing applied-research or research-engineering teams, deep credibility in RL/post-training/agents/evaluation/ML systems, exceptional recruiting judgment, and experience with demanding external technical stakeholders.
Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.
Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.
Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.
Evidence of exceptional work
We value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.
Cash Compensation Range of $150-300k + equity incentives
Full-time, in-person work from our New York City office.
Visa Sponsorship & relocation support
Professional Development budget
Team Off-sites & conference attendance
If you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n\nROLE IMPACT\n\nWe are looking for a senior regional player-coach who can establish that group through personally led research and field programs. You will begin close to the work: shape consequential external problems into coherent research programs, lead the hardest early experiments and deployments, and recruit exceptional researchers who want the same combination of frontier methods and real-world responsibility. As the group grows, you will set its research direction and the technical and experimental standards that let it move quickly without losing rigor.\n\n\nRESPONSIBILITIES\n\nApplied Research at Prime\n\nWe turn a workflow into tasks, environments, evaluations, and reward signals, run post-training experiments against real baselines, and deploy what works. When a result looks good, we find out whether the gain came from the model or from the data, harness, inference, or infrastructure. The reusable parts become shared research, open-source systems, and product capabilities.\n\nEvery Applied Research role is externally facing. Most researchers work directly with customers, users, or research partners while maintaining a substantive research agenda.\n\nRole scope\n\nThis is a New York leadership role within Prime's broader Applied Research function. It requires research taste, technical depth, customer and partner judgment, exceptional recruiting, and the willingness to remain a builder. You will stay directly involved with external program teams while leading technical work, recruiting, and developing the New York group.\n\nRepresentative early problems\n\n - Turn the first portfolio of New York field programs into research agendas with strong owners, comparable evidence, and explicit paths from domain result to shared capability.\n\n - Personally lead a high-stakes program through environment design, baseline, post-training, evaluation, and deployment while establishing the quality bar for the growing team.\n\n - Recruit the founding New York researchers and establish a working rhythm that keeps review rigorous without slowing the team.\n\nWhat you will own\n\n - A small number of the hardest New York partner and customer research programs.\n\n - Recruiting and development of the founding New York Applied Research group.\n\n - Technical standards for environments, evaluations, post-training experiments, failure analysis, and deployed results.\n\n - Meaningful personal research, cross-program technical direction, and decisions about what becomes shared research or product capability as the team grows.\n\n\nREQUIREMENTS\n\n - A track record leading high-performing applied-research or research-engineering teams, deep credibility in RL/post-training/agents/evaluation/ML systems, exceptional recruiting judgment, and experience with demanding external technical stakeholders.\n\n - Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.\n\n - Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and a willingness to change direction when the evidence changes.\n\n - Strong Python and the ability to build artifacts that other researchers and engineers can run, inspect, and extend.\n\n - Comfort owning ambiguous problems and keeping partners and teammates current as you resolve them.\n\nEvidence of exceptional work\n\nWe value publications, open-source systems, benchmarks, and shipped products as evidence of exceptional work. Show us something difficult you built, what failed, and how the evidence changed your approach.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-300k + equity incentives\n\n - Full-time, in-person work from our New York City office.\n\n - Visa Sponsorship & relocation support\n\n - Professional Development budget\n\n - Team Off-sites & conference attendance\n\nIf you're excited about the opportunity to build the foundation for the future of decentralized AI and create a platform that empowers developers and researchers to push the boundaries of what's possible, we'd love to hear from you."},{"id":"3e11ec2c-34c1-45c8-a732-ab3001225f25","title":"Senior Technical Partnerships Manager","department":"Growth","team":"Growth","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-10-08T18:46:20.266+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/3e11ec2c-34c1-45c8-a732-ab3001225f25","applyUrl":"https://jobs.ashbyhq.com/primeintellect/3e11ec2c-34c1-45c8-a732-ab3001225f25/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
No one builds the AI infrastructure stack alone. The GPU makers, cloud and datacenter providers, model labs, and enterprise platforms we work with are core to how we reach customers and how we ship. The partnerships that matter in this industry are not logos on a slide — they are shipped integrations, field teams that can actually sell and support a joint solution, and customers who win because both companies showed up together. Very few people know how to make those partnerships real at the technical and commercial level simultaneously.
You will own our most strategic technical alliances end to end: NVIDIA, hyperscalers, neoclouds, datacenter operators, and the next wave of partners across models, tooling, and enterprise. You will be the person a partner's engineering, product, and field teams trust to turn a signed MOU into something that moves pipeline and ships product.
The work sits at the intersection of technical depth and commercial execution. You will go deep enough on GPUs, networking, distributed training, and inference stacks to hold your own with partner engineers, and broad enough to run joint go-to-market, executive relationships, and a scalable partner program. You will scope proofs of concept with joint customers, build the enablement that lets partner sellers position us, secure marketplace and co-sell motions, and bring partner roadmaps back into our product and research decisions.
The partnerships you build will directly shape how Prime Intellect reaches the labs, enterprises, and research teams building on our stack — and how fast we can deliver compute, training, and environments to them.
Partner Relationship Management
Own 3–5 tier-1 technical partnerships (e.g., NVIDIA, hyperscalers, neoclouds, datacenter operators) as the single accountable lead
Build relationships across the partner org at every level: alliance managers, developer relations, product and engineering, field sales, and executives
Run quarterly business reviews and set joint annual plans with clear goals, owners, and metrics
Navigate partner programs (e.g., NVIDIA Inception, NPN, cloud marketplace and co-sell programs) to unlock credits, benefits, tiering, and visibility
Negotiate and manage partnership agreements, MOUs, and joint commitments with Legal and Finance
Joint Customer Engagement
Lead and join joint customer calls with partner account teams, from discovery through technical deep dives and close
Act as the technical bridge in deals: map customer workloads to the joint solution, scope proofs of concept, and remove blockers
Coordinate escalations across both companies when joint customers hit issues
Build a pipeline of partner-sourced and partner-influenced opportunities, and report on it
Technical Integration & Product Feedback
Drive joint technical work: integrations, benchmarks, validated configurations, certifications, and optimizations on partner hardware and software
Get early access to partner roadmaps, hardware, and SDKs, and translate them into product opportunities for us
Bring partner and customer feedback into our product and research roadmaps, and advocate for our needs in the partner's roadmap
Write or co-author reference architectures, technical blog posts, and solution briefs
Training & Enablement
Build enablement content for partner field teams: pitch decks, battlecards, demo scripts, reference architectures, and FAQs
Deliver live and recorded training to partner sellers, solution architects, and support engineers so they can position, sell, and troubleshoot our products
Enable our own team on partner products, roadmaps, and programs
Track enablement reach and impact: people trained, certified, and actively co-selling
Joint Go-to-Market & Marketing
Plan co-marketing: launches, case studies, webinars, joint press, and partner showcase placements
Represent Prime Intellect at partner events and industry conferences (e.g., NVIDIA GTC), including talks, booths, and customer meetings
Secure marketplace listings and co-sell motions with cloud and channel partners
Partner Program & Strategy
Identify and evaluate new strategic partners across compute, models, tooling, and enterprise
Design a scalable partner program: tiers, benefits, onboarding, and playbooks
Own partnership metrics and report regularly to leadership on revenue impact, pipeline, and technical milestones
7+ years in technical partnerships, alliances, solutions architecture, or technical business development, with at least 3 years managing strategic partners
Direct experience working with NVIDIA, a hyperscaler (AWS, GCP, Azure), or a GPU cloud provider — ideally on the partner side of that relationship
Strong technical grounding in AI infrastructure: GPUs and accelerators, networking (InfiniBand, RoCE), distributed training and inference, and the CUDA and PyTorch ecosystem
A track record of co-selling — you've run joint customer calls and closed or influenced deals alongside partner field teams
Experience building and delivering technical enablement to sales and engineering audiences
Comfortable presenting to executives, engineers, and conference audiences
Able to run a complex, multi-stakeholder relationship with little oversight, and to say no when a joint initiative isn't worth it
Clear, concise writer: plans, briefs, and technical docs
High ownership — you see gaps and build the fix before anyone asks
AI-native in how you work: you use LLMs, automation, and programmatic tools to move faster
Bonus:
Prior experience inside NVIDIA's partner or developer ecosystem, or inside a hyperscaler's AI/ML partner organization
Hands-on background as a solutions architect or ML engineer before moving into partnerships
Experience standing up a partner program from scratch at an early-stage infrastructure company
Existing relationships with alliance, field, and engineering teams at major compute providers
The AI infrastructure ecosystem is being assembled right now, and the partnerships that define it are still being written. You'll own the relationships that determine how Prime Intellect reaches the market, work directly with leadership and the engineering team on what we build and with whom, and sit in the room with partner executives and engineers at the companies shaping the next decade of compute. The network and expertise you build here are increasingly rare and valuable across the industry.
Cash compensation range of $250–300k + meaningful equity
San Francisco or New York, with travel to partner sites, customer meetings, and conferences (about 25–30%)
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
We are a team of about 60, split between our San Francisco office and remote staff, and we are hiring quickly. Every researcher, engineer, and operator who joins us has a first day, a first week, and a first impression of how seriously we take the people who build here. Right now, that experience depends on a lot of people doing a lot of small things right. This role makes it a system.
You will own the employee lifecycle from the moment a candidate is scheduled for an interview through a new hire's first weeks. You will design every new hire's first day, run onboarding end to end, coordinate recruiting logistics across multiple open searches, and help bring the team together in person. You will also be the person who looks at how this function works and makes it faster, more reliable, and more personal — including where AI tools and automation can take the repetitive parts off your plate so you can spend your time on the parts that only a person can do.
This is an in-person role based in our San Francisco office, reporting to our Executive Business Partner and working closely with the Head of Talent, hiring managers, IT, and finance.
Onboarding & Employee Experience
Design the first-day schedule and experience for every new hire so they feel welcomed and productive from day one, with a white-glove approach for executive hires
Set up desks and equipment before day one: laptop, monitor, welcome kit in place; coordinate provisioning, accounts, and tool access with IT
Coordinate finance onboarding — payroll, benefits enrollment, expense tools, company card — and office access: door access, building entry, and meal program
Complete required onboarding steps accurately and on time: background checks, I-9 verification, employee records, and HRIS setup
Identify immigration and visa needs early, route them to the right people, and track them to resolution
Maintain and continuously improve the onboarding system so every hire gets a consistent, well-run first week
Support offboarding: recover equipment and remove access
Recruiting Coordination
Own interview scheduling across multiple open roles so every candidate and interviewer shows up informed, prepared, and on time
Book travel and host candidates for onsite interviews and work trials, creating a warm, human candidate experience
Keep candidates and hiring teams informed throughout the process with timely, clear communication
Maintain Ashby as the single source of truth: stage updates, candidate notes, and pipeline hygiene
Own scheduling timelines and SLAs, surface risks early, and keep recruiters and hiring managers in the loop
Systems, Culture & Team Operations
Find smarter, faster ways to run this function, including automating repetitive work with AI tools
Track open hiring and onboarding tasks and follow through on every commitment
Contribute to an in-person culture that builds connection across in-office and remote teams
Help plan and run team onsites and offsites, including attendee logistics and travel
Take on recruiting operations and sourcing projects as the function grows
New hires feel expected and cared for. Before day one, they know where to go, what to expect, and who to ask for help.
Laptops arrive on time, accounts work, schedules are clear, and first-week travel is arranged. You check directly with each new hire rather than assuming everything is set.
Compliance steps — background checks, I-9s, records — are complete and accurate without anyone having to chase them.
Recruiting, IT, managers, and office teams know what they own. You spot delays and resolve them before they affect a start date.
Trackers, checklists, and communications stay current as hiring volume grows.
2+ years in recruiting coordination, people operations, HR, or a similar high-volume coordination role
Proactive coordination: you can juggle many starts, owners, and deadlines without losing the details, and you follow up before anyone has to ask
Real care for the employee and candidate experience — a warm welcome is built from both thoughtful communication and reliable execution
Comfort managing background checks, I-9 verification, employee records, and HRIS setup, and handling sensitive information with discretion, including for senior and executive candidates
Cross-functional follow-through: you make ownership clear and stay with an issue until it's resolved
Process improvement instinct: you spot recurring gaps in a checklist, handoff, or message and fix them
Highly skilled in tools like Google Calendar, Slack, and Notion, with demonstrated ability to handle high-volume work efficiently
AI-native in how you work: you use LLMs and automation to take repetitive work off your plate
Low ego, high ownership — as willing to improve a workflow as to track down a delayed laptop
Able to work in person in our San Francisco office
Bonus:
Experience with Ashby or another modern applicant tracking system
Experience running onboarding at a fast-growing startup
Experience coordinating immigration or visa processes with outside counsel
You will shape how every person who joins Prime Intellect experiences their first weeks, at a company where the people being onboarded are among the best researchers and engineers in the world. You will build the people operations foundation from an early stage, work directly with leadership and the Head of Talent, and have real room to grow into recruiting operations, people operations, or wherever your strengths take you as the company scales.
$90,000–$120,000 base salary + equity
In-person role in our San Francisco office (501 Folsom St)
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
We train open frontier models and ship the same stack to our customers. Its spans the full stack of training, deploying and continuously improving models — compute, large-scale RL, environments, sandboxes, evals, and deployment.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
The next frontier in AI will not be unlocked by models alone. It will be unlocked by systems that let those models train faster, adapt continuously, and operate across real environments at scale.
That infrastructure does not exist yet in the form the world needs.
We’re building it.
Build and optimize the systems infrastructure behind large-scale RL and distributed training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.
Improve end-to-end training efficiency across compute, memory, networking, and scheduling layers.
Design and implement low-level performance optimizations, including kernels, communication paths, and runtime improvements.
Work on distributed training systems spanning data, tensor, and pipeline parallel workloads.
Help shape the architecture of our RL training stack, including async rollout and post-training systems.
Contribute to open-source libraries and internal infrastructure used for frontier-scale model training.
Collaborate closely with researchers and infrastructure engineers to translate bottlenecks into concrete systems improvements.
Stay at the frontier of training systems, inference systems, compiler/runtime tooling, and hardware-aware optimization techniques.
Strong systems engineering experience in AI/ML infrastructure, especially around large-scale model training or inference.
Deep familiarity with PyTorch and distributed training frameworks such as PyTorch Distributed, DeepSpeed, FSDP, Megatron, vLLM, Ray, or related tooling.
Experience optimizing training performance across kernels, memory movement, communication overhead, or parallelization strategy.
Hands-on experience with large-scale training techniques including data parallelism, tensor parallelism, and pipeline parallelism.
Strong understanding of GPU architecture, profiling, and performance debugging.
Ability to identify bottlenecks across the stack and drive improvements from first principles.
Comfort working in a fast-moving environment with ambiguous problems and high ownership.
Especially Exciting (optional)
Experience writing or optimizing CUDA / Triton kernels.
Experience with compiler or runtime optimization for ML systems.
Experience working on RL training infrastructure, rollout systems, or asynchronous training pipelines.
Experience with multi-node GPU clusters and high-performance networking.
Contributions to open-source ML systems or infrastructure projects.
Interest in publishing technical work or sharing insights through engineering blogs and technical writing.
Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.
Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.
Visa sponsorship and relocation assistance for international candidates.
Quarterly team off-sites, hackathons, conferences and learning opportunities.
Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.
If you’re excited about building the systems foundation for frontier-scale RL and open superintelligence, we’d love to hear from you.
","descriptionPlain":"OWN YOUR INTELLIGENCE\n\n \n\nPrime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.\n\n \n\nOur platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.\n\n \n\nWe train open frontier models and ship the same stack to our customers. Its spans the full stack of training, deploying and continuously improving models — compute, large-scale RL, environments, sandboxes, evals, and deployment.\n\nPrime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.\n\n \n\n\nROLE IMPACT\n\nThe next frontier in AI will not be unlocked by models alone. It will be unlocked by systems that let those models train faster, adapt continuously, and operate across real environments at scale.\n\nThat infrastructure does not exist yet in the form the world needs.\n\nWe’re building it.\n\n\nRESPONSIBILITIES\n\n - Build and optimize the systems infrastructure behind large-scale RL and distributed training workloads by contributing to our prime-rl https://github.com/PrimeIntellect-ai/prime-rl framework.\n\n - Improve end-to-end training efficiency across compute, memory, networking, and scheduling layers.\n\n - Design and implement low-level performance optimizations, including kernels, communication paths, and runtime improvements.\n\n - Work on distributed training systems spanning data, tensor, and pipeline parallel workloads.\n\n - Help shape the architecture of our RL training stack, including async rollout and post-training systems.\n\n - Contribute to open-source libraries and internal infrastructure used for frontier-scale model training.\n\n - Collaborate closely with researchers and infrastructure engineers to translate bottlenecks into concrete systems improvements.\n\n - Stay at the frontier of training systems, inference systems, compiler/runtime tooling, and hardware-aware optimization techniques.\n\n\nREQUIREMENTS\n\n - Strong systems engineering experience in AI/ML infrastructure, especially around large-scale model training or inference.\n\n - Deep familiarity with PyTorch and distributed training frameworks such as PyTorch Distributed, DeepSpeed, FSDP, Megatron, vLLM, Ray, or related tooling.\n\n - Experience optimizing training performance across kernels, memory movement, communication overhead, or parallelization strategy.\n\n - Hands-on experience with large-scale training techniques including data parallelism, tensor parallelism, and pipeline parallelism.\n\n - Strong understanding of GPU architecture, profiling, and performance debugging.\n\n - Ability to identify bottlenecks across the stack and drive improvements from first principles.\n\n - Comfort working in a fast-moving environment with ambiguous problems and high ownership.\n\nEspecially Exciting (optional)\n\n - Experience writing or optimizing CUDA / Triton kernels.\n\n - Experience with compiler or runtime optimization for ML systems.\n\n - Experience working on RL training infrastructure, rollout systems, or asynchronous training pipelines.\n\n - Experience with multi-node GPU clusters and high-performance networking.\n\n - Contributions to open-source ML systems or infrastructure projects.\n\n - Interest in publishing technical work or sharing insights through engineering blogs and technical writing.\n\n\nBENEFITS AND PERKS\n\n - Cash Compensation Range of $150-350k, including equity incentives, aligning your success with the growth and impact of Prime Intellect.\n\n - Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco.\n\n - Visa sponsorship and relocation assistance for international candidates.\n\n - Quarterly team off-sites, hackathons, conferences and learning opportunities.\n\n - Opportunity to work with a talented, hard-working and mission-driven team, united by a shared passion for leveraging technology to accelerate science and AI.\n\nIf you’re excited about building the systems foundation for frontier-scale RL and open superintelligence, we’d love to hear from you."},{"id":"d55d1904-1c78-4d87-b71e-26df5c700e8d","title":"Product Marketing Lead","department":"Growth","team":"Growth","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-10-14T18:16:08.680+00:00","isListed":true,"isRemote":null,"workplaceType":null,"address":{"postalAddress":{"addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/primeintellect/d55d1904-1c78-4d87-b71e-26df5c700e8d","applyUrl":"https://jobs.ashbyhq.com/primeintellect/d55d1904-1c78-4d87-b71e-26df5c700e8d/application","descriptionHtml":"Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, brings together compute, environments, evaluations, sandboxes, training, and deployment so teams can build and improve models around their own workflows. We build open frontier models ourselves and make the infrastructure behind that work available to other AI teams. Our ambition is to give every ambitious team the ability to own its intelligence.
We're looking for people who want to work close to frontier research, build real infrastructure, and help shape a market that's still taking form.
As our first Product Marketing Lead, you'll fully own the zero to one build of the product marketing function at Prime Intellect. You'll help define how customers understand what we're building, why it matters, and where it fits in their work. That means shaping a new category while making the value of our products concrete enough for a customer to act on today.
Our primary commercial motion is sales-led. You'll work directly with customers, founders, Sales, Product, Applied Research, and Engineering to find the use cases that resonate, build the sales narrative, and turn early customer results into proof. You'll own positioning and messaging as a living body of work, testing it in real conversations and changing it as we learn.
This is a hands-on role on a small team. You'll set direction and write the first version yourself, from a product page to a launch story to the deck for tomorrow's customer meeting.
Define and evolve our product positioning and category narrative. Help customers understand what open post-training infrastructure makes possible, why they need it, and why Prime Intellect is the right partner.
Turn technical capabilities into clear messaging for developers, AI teams, and the people making buying decisions. Build the core narrative for Prime Intellect's Open SuperIntelligence Stack and the Model Factory offering via product pages and use-case stories that connect our long-term ambition to problems customers have today.
Build a regular practice of customer interviews, sales-call participation, and win/loss analysis. Learn how customers describe their problems, what triggers a buying decision, and where our story falls short.
Test and revise messaging based on what you hear and what moves opportunities forward. Bring those insights to Product, Engineering and Applied Research to sharpen target use cases, inform packaging, and help distinguish repeatable needs from one-off requests.
Own product launches from positioning and audience selection through launch content, seller readiness, and follow-through. Work with Marketing and Sales to turn each launch into relevant customer conversations and qualified opportunities.
Build the materials sellers need to win: pitch decks, demos with technical partners, competitive guidance, objection handling, and customer proof. Stay close to live opportunities, join calls when useful, and keep improving the materials based on how customers respond.
Turn early customer work into case studies, references, and clear accounts of what changed for the customer. Build design-partner relationships that strengthen our story and feed learning back into the product.
Establish the first product marketing processes and priorities. Measure whether our work helps customers understand the product, progress through evaluation, buy, and expand their use. Use AI tools and agents to accelerate research, synthesis, and production while keeping a high bar for accuracy and craft.
8+ years of product marketing or closely related experience, including substantial experience marketing technical products built for developers, engineers, or AI teams.
Experience taking a product or product marketing function from zero to one. You've built positioning, launches, and sales materials when the audience, category, or playbook was still being defined.
A track record supporting a sales-led motion. You've worked directly with sellers and technical partners on real opportunities, and can point to messaging and materials that helped customers make a buying decision.
Strong customer research instincts. You can show how customer conversations changed your positioning, challenged an assumption, or led you to focus on a different use case.
Technical fluency and curiosity. You can learn a complex product, explain how it works, and earn credibility with developers and researchers while making its value clear to a business buyer.
Excellent writing and strong editorial judgment. You can turn a dense technical discussion into a clear story without losing what makes the product distinctive.
Demonstrated, hands-on use of AI tools and agents in your work. We'll ask you to show us how you work.
Comfort with ambiguity and a bias toward shipping. You can choose what matters, produce a strong first version, and improve it as evidence comes in.
Experience with AI/ML infrastructure, post-training, reinforcement learning, evaluations, or agent platforms.
Experience helping define an emerging category or repositioning a technical product as the product and market evolved.
A technical background, hands-on product experience, or experience building a customer advisory or design-partner program.
Competitive cash compensation and meaningful equity
Visa sponsorship and relocation support
Professional development budget
Team off-sites and conference attendance
The opportunity to help shape the market for open AI infrastructure
Prime Intellect is building the open superintelligence stack: the infrastructure frontier AI labs build internally, made available to every ambitious AI team.
Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment into one full-stack system for post-training at frontier scale - from SFT and RL to tool use, agent workflows, and continuously improving production models. We are building open frontier AI: open-source models trained end to end for long-horizon tasks like autonomous research, and the full-stack platform our own research team uses to build them. The next generation of AI companies, enterprises, and research teams do not just need more GPUs. They need the ability to turn their own workflows, tools, data, and feedback loops into superintelligence they own.
Prime Intellect has raised $150M in total funding from Founders Fund, Radical Ventures, NVIDIA, and exceptional AI, infrastructure, and enterprise operators — including Andrej Karpathy, Dwarkesh Patel, and leaders and founders from Ramp, Perplexity, Harvey, Mercor, Zapier, Datadog, Cognition, OpenAI, Thinking Machines, Together AI, SemiAnalysis, LangChain, Browserbase, Cloudflare, Sierra, Databricks, Airbnb, OpenRouter, Standard Intelligence, Fleet, Core Auto, and more. We are looking for people who want to build at the intersection of frontier research, real infrastructure, and go-to-market for a category that does not fully exist yet.
We are a team of about 60, headquartered in San Francisco, with teammates around the world. The office is where a globally distributed team becomes one team: where researchers and engineers work side by side, where customers and partners visit, and where onsites and offsites bring everyone together. Making that space work well, every day, is not a background task — it shapes how the company feels to work at.
As Office Manager, you will run our SF office day to day and help plan the in-person moments that bring the team together: onsites, offsites, team dinners, and events in San Francisco, New York, and Berlin. You will be office manager, host, planner, and fixer in one. The first big project is our office move, which you will help drive end to end.
This is a high-trust, high-autonomy role reporting to our Executive Business Partner, fully onsite five days a week. If you like variety, take pride in a space that runs well, and spot what's broken before anyone asks, you'll fit right in.
Office Operations
Help drive the office move end to end — vendors, logistics, and onsite needs — as the first priority
Run SF HQ day to day: space, kitchen and snacks, furniture, supplies, mail, and packages
Do daily walkthroughs to keep the office clean, safe, stocked, and compliant with health and safety requirements
Serve as the point of contact for building management, maintenance, and facilities vendors, including cleaning and AV
Manage access control and visitor registration, including badges and guest credentials
Be the onsite point person for equipment, printers, monitors, and conference room tech, in partnership with IT
Maintain desk assignments, meeting room bookings, and phone booths
Support purchasing and vendor invoices with the Executive Business Partner
Onsites, Offsites & Events
Help plan and run company onsites and offsites: venues, housing, travel, catering, group activities, and daily onsite operations
Support logistics for executive leadership offsites, conferences, and customer visits
Plan happy hours, dinners, celebrations, and events in SF and other cities, including New York and Berlin
Manage event vendors alongside the Executive Business Partner
Travel & Executive Support
Book and coordinate team travel for offsites, events, and in-office visits
Support international travelers with logistics such as ESTA and visa invitation letters
Back up the Executive Business Partner during PTO and peak periods: calendars, travel, expenses, and ad-hoc needs
Workplace Experience
Host guests and shape the in-person experience for visitors, candidates, and customers
Run employee experience programs, track team birthdays, and help build culture through team events
Help manage the Shopify store and merch
3+ years in office management, workplace operations, hospitality, or event coordination, ideally at a fast-growing startup
Hands-on experience planning multi-day offsites or events for 30+ people
Fluency with Google Workspace, Slack, and Notion; comfortable with laptops, docks, and conference room tech
Ridiculously organized, with strong follow-through and attention to detail
Warm, unflappable, and great with people at every level
Comfortable with shifting priorities and making the call when no one else is around to make it
Sound judgment with confidential information
AI-native in how you work: you use LLMs and automation to take repetitive work off your plate
Based in SF and able to be fully onsite five days a week; able to lift up to 25 lbs
Bonus:
International travel and group housing logistics
Experience with access control or visitor management systems
Experience managing an office move or build-out
Experience with Shopify
You will own how Prime Intellect works in person — the office, the events, the moments that turn a distributed team into one — at a company where the people walking through the door are among the best researchers and engineers in the world. You will work directly with leadership on a small team with no layers between you and the work, and have real room to grow into workplace operations, events, or people operations as the company scales.
$95,000–$120,000 base salary + equity
Fully onsite role at our San Francisco HQ (501 Folsom St, SoMa), with occasional travel for offsites and events
Benefits included
Professional development budget
Team off-sites and conferences
A front-row seat to building the infrastructure layer for open AI