We built Autolabel because access to clean, labeled data is a huge bottleneck for most ML/data science teams. The most capable LLMs are able to label data with high accuracy, and at a fraction of the cost and time compared to manual labeling. With Autolabel, you can leverage LLMs to label any text dataset with <5 lines of code.
We\u2019re eager for your feedback!"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Show HN: Autolabel, a Python library to label and enrich text data with LLMs"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://github.com/refuel-ai/autolabel"}},"_tags":["story","author_nihit-desai","story_36409201","show_hn"],"author":"nihit-desai","children":[36409582,36410063,36410705,36411433,36414619,36415147,36417025,36424463],"created_at":"2023-06-20T19:26:02Z","created_at_i":1687289162,"num_comments":22,"objectID":"36409201","points":153,"story_id":36409201,"story_text":"Hi HN! I'm excited to share Autolabel, an open-source Python library to label and enrich text datasets with any Large Language Model (LLM) of your choice.
We built Autolabel because access to clean, labeled data is a huge bottleneck for most ML/data science teams. The most capable LLMs are able to label data with high accuracy, and at a fraction of the cost and time compared to manual labeling. With Autolabel, you can leverage LLMs to label any text dataset with <5 lines of code.
We\u2019re eager for your feedback!","title":"Show HN: Autolabel, a Python library to label and enrich text data with LLMs","updated_at":"2024-09-20T14:22:12Z","url":"https://github.com/refuel-ai/autolabel"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mhb"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hands off \u2013 Garmin develops autoland system"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.aopa.org/News-and-Media/All-News/2020/January/Pilot/hands-off"}},"_tags":["story","author_mhb","story_21409114"],"author":"mhb","created_at":"2019-10-31T14:23:37Z","created_at_i":1572531817,"num_comments":0,"objectID":"21409114","points":11,"story_id":21409114,"title":"Hands off \u2013 Garmin develops autoland system","updated_at":"2024-09-20T05:06:05Z","url":"https://www.aopa.org/News-and-Media/All-News/2020/January/Pilot/hands-off"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"msvan"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Show HN: Autoliner \u2013 write a bot to control a virtual airline"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autoliner.app/"}},"_tags":["story","author_msvan","story_46857754","show_hn"],"author":"msvan","created_at":"2026-02-02T16:18:07Z","created_at_i":1770049087,"num_comments":0,"objectID":"46857754","points":5,"story_id":46857754,"title":"Show HN: Autoliner \u2013 write a bot to control a virtual airline","updated_at":"2026-03-05T23:29:19Z","url":"https://autoliner.app/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"aschobel"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin announces autonomous emergency autoland for general aviation"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.garmin.com/en-US/autonomi/"}},"_tags":["story","author_aschobel","story_21398066"],"author":"aschobel","created_at":"2019-10-30T14:23:55Z","created_at_i":1572445435,"num_comments":0,"objectID":"21398066","points":5,"story_id":21398066,"title":"Garmin announces autonomous emergency autoland for general aviation","updated_at":"2024-09-20T05:13:45Z","url":"https://www.garmin.com/en-US/autonomi/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"msvan"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Show HN: Autoliner \u2013 write a bot to control a virtual airline"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autoliner.app/"}},"_tags":["story","author_msvan","story_46875162","show_hn"],"author":"msvan","children":[46895363],"created_at":"2026-02-03T18:35:27Z","created_at_i":1770143727,"num_comments":1,"objectID":"46875162","points":4,"story_id":46875162,"title":"Show HN: Autoliner \u2013 write a bot to control a virtual airline","updated_at":"2026-03-05T23:30:33Z","url":"https://autoliner.app/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alangibson"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Emergency Autoland Deployed for the first time"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"}},"_tags":["story","author_alangibson","story_46364322"],"author":"alangibson","children":[46396610],"created_at":"2025-12-23T11:03:30Z","created_at_i":1766487810,"num_comments":1,"objectID":"46364322","points":4,"story_id":46364322,"title":"Garmin Emergency Autoland Deployed for the first time","updated_at":"2026-03-05T23:16:59Z","url":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jumbopapa"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin announces the first Autoland system for general aviation aircraft"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.businesswire.com/news/home/20191030005354/en/"}},"_tags":["story","author_jumbopapa","story_21397177"],"author":"jumbopapa","children":[21397329],"created_at":"2019-10-30T12:58:26Z","created_at_i":1572440306,"num_comments":1,"objectID":"21397177","points":4,"story_id":21397177,"title":"Garmin announces the first Autoland system for general aviation aircraft","updated_at":"2024-09-20T05:13:39Z","url":"https://www.businesswire.com/news/home/20191030005354/en/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rmason"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Plane makes safe emergency landing with first use of Autoland system (2M video)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=XawXSLRCM3o"}},"_tags":["story","author_rmason","story_46369041"],"author":"rmason","children":[46369054,46369093,46379213],"created_at":"2025-12-23T20:18:37Z","created_at_i":1766521117,"num_comments":4,"objectID":"46369041","points":3,"story_id":46369041,"title":"Plane makes safe emergency landing with first use of Autoland system (2M video)","updated_at":"2026-03-05T23:17:20Z","url":"https://www.youtube.com/watch?v=XawXSLRCM3o"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"r2sk5t"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Is Automating In-Flight Emergencies (Autoland, Smart Glide, ESP)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.garmin.com/en-US/newsroom/press-release/aviation/garmin-unveils-g2000-prime-integrated-flight-deck/"}},"_tags":["story","author_r2sk5t","story_48996106"],"author":"r2sk5t","children":[48996107],"created_at":"2026-07-21T18:23:17Z","created_at_i":1784658197,"num_comments":0,"objectID":"48996106","points":3,"story_id":48996106,"title":"Garmin Is Automating In-Flight Emergencies (Autoland, Smart Glide, ESP)","updated_at":"2026-07-21T20:52:57Z","url":"https://www.garmin.com/en-US/newsroom/press-release/aviation/garmin-unveils-g2000-prime-integrated-flight-deck/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tjohns"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Emergency Autoland Flight Trial"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=d-ruFmgTpqA"}},"_tags":["story","author_tjohns","story_21405699"],"author":"tjohns","created_at":"2019-10-31T04:14:15Z","created_at_i":1572495255,"num_comments":0,"objectID":"21405699","points":3,"story_id":21405699,"title":"Garmin Emergency Autoland Flight Trial","updated_at":"2024-09-20T05:05:36Z","url":"https://www.youtube.com/watch?v=d-ruFmgTpqA"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dejavucoder"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"AutoLatex: LLM-powered text/image to LaTeX with live rendering Chrome extension"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://chromewebstore.google.com/detail/autolatex/fmjogafcialngfcopflagalflkcnlebk"}},"_tags":["story","author_dejavucoder","story_41310289"],"author":"dejavucoder","children":[41310290],"created_at":"2024-08-21T13:50:34Z","created_at_i":1724248234,"num_comments":1,"objectID":"41310289","points":2,"story_id":41310289,"title":"AutoLatex: LLM-powered text/image to LaTeX with live rendering Chrome extension","updated_at":"2024-09-20T17:35:43Z","url":"https://chromewebstore.google.com/detail/autolatex/fmjogafcialngfcopflagalflkcnlebk"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"woliveirajr"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Today a new "feature" appeared in Gmail: A "autolabel" button. And it displaces other icons as it appears/disappears "randomly". Is it another step from Google to the shitfication of things ?
Link: https://imgur.com/a/IAdrItn"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Ask HN: \"Autolabel\" in Gmail?"}},"_tags":["story","author_woliveirajr","story_40254071","ask_hn"],"author":"woliveirajr","children":[40254129],"created_at":"2024-05-04T01:29:15Z","created_at_i":1714786155,"num_comments":1,"objectID":"40254071","points":2,"story_id":40254071,"story_text":"Today a new "feature" appeared in Gmail: A "autolabel" button. And it displaces other icons as it appears/disappears "randomly". Is it another step from Google to the shitfication of things ?
Link: https://imgur.com/a/IAdrItn","title":"Ask HN: \"Autolabel\" in Gmail?","updated_at":"2024-09-20T17:01:17Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mhazelwood"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Develops Autoland System"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.aopa.org/news-and-media/all-news/2020/january/pilot/hands-off"}},"_tags":["story","author_mhazelwood","story_21417695"],"author":"mhazelwood","children":[21417716],"created_at":"2019-11-01T11:24:41Z","created_at_i":1572607481,"num_comments":1,"objectID":"21417695","points":2,"story_id":21417695,"title":"Garmin Develops Autoland System","updated_at":"2024-09-20T05:06:58Z","url":"https://www.aopa.org/news-and-media/all-news/2020/january/pilot/hands-off"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hoansdz"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hi HN, I\u2019m a student and I\u2019ve spent the last few months building Autolang\u2014an embedded language and VM from scratch in C++.
I built it because I needed a specialized runtime for high-frequency, short-lived tasks (like AI agent loops) where Python/Lua startup latency and GC overhead become actual bottlenecks.
Core Architecture:
Arena-restart: No Tracing GC. It prioritizes allocation speed (~2x faster than Lua). Memory is wiped instantly between tasks via a restart() mechanism.
Fast Compilation: 100k classes compile in ~888ms (Single-pass).
Lightweight: The VM is designed for a ~2ms startup target.
The Trade-off: Execution is currently 2x-5x slower than Lua as I'm still optimizing the bytecode dispatch (if-chains).
I\u2019d love to hear your thoughts on using Arena-restart for autonomous scripts and any advice on optimizing dispatch without bloating the startup time."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Show HN: Autolang-A C++ VM with 2ms startup time and arena-restart memory model"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autolang.vercel.app/docs/introduction"}},"_tags":["story","author_hoansdz","story_47206397","show_hn"],"author":"hoansdz","created_at":"2026-03-01T13:11:47Z","created_at_i":1772370707,"num_comments":0,"objectID":"47206397","points":2,"story_id":47206397,"story_text":"Hi HN, I\u2019m a student and I\u2019ve spent the last few months building Autolang\u2014an embedded language and VM from scratch in C++.
I built it because I needed a specialized runtime for high-frequency, short-lived tasks (like AI agent loops) where Python/Lua startup latency and GC overhead become actual bottlenecks.
Core Architecture:
Arena-restart: No Tracing GC. It prioritizes allocation speed (~2x faster than Lua). Memory is wiped instantly between tasks via a restart() mechanism.
Fast Compilation: 100k classes compile in ~888ms (Single-pass).
Lightweight: The VM is designed for a ~2ms startup target.
The Trade-off: Execution is currently 2x-5x slower than Lua as I'm still optimizing the bytecode dispatch (if-chains).
I\u2019d love to hear your thoughts on using Arena-restart for autonomous scripts and any advice on optimizing dispatch without bloating the startup time.","title":"Show HN: Autolang-A C++ VM with 2ms startup time and arena-restart memory model","updated_at":"2026-03-05T23:38:58Z","url":"https://autolang.vercel.app/docs/introduction"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"taubek"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"The accountability problem exposed by the first Garmin Autoland deployment"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://theaircurrent.com/aviation-safety/garmin-autoland-first-deployment-accountability-pilots-automation/"}},"_tags":["story","author_taubek","story_46793561"],"author":"taubek","created_at":"2026-01-28T10:32:52Z","created_at_i":1769596372,"num_comments":0,"objectID":"46793561","points":2,"story_id":46793561,"title":"The accountability problem exposed by the first Garmin Autoland deployment","updated_at":"2026-03-05T23:27:33Z","url":"https://theaircurrent.com/aviation-safety/garmin-autoland-first-deployment-accountability-pilots-automation/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bookofjoe"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Autoland: Passenger-activated emergency landing system"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://youtu.be/YTUk7-wmfCs"}},"_tags":["story","author_bookofjoe","story_21796724"],"author":"bookofjoe","children":[21796747,21796915],"created_at":"2019-12-15T15:41:47Z","created_at_i":1576424507,"num_comments":2,"objectID":"21796724","points":1,"story_id":21796724,"title":"Garmin Autoland: Passenger-activated emergency landing system","updated_at":"2024-09-20T05:24:42Z","url":"https://youtu.be/YTUk7-wmfCs"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jaredwiener"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin Emergency Autoland Deployed For the First Time"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"}},"_tags":["story","author_jaredwiener","story_46396604"],"author":"jaredwiener","children":[46396612],"created_at":"2025-12-26T21:42:20Z","created_at_i":1766785340,"num_comments":1,"objectID":"46396604","points":1,"story_id":46396604,"title":"Garmin Emergency Autoland Deployed For the First Time","updated_at":"2026-03-05T23:19:06Z","url":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"strangattractor"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Emergency Autoland deployed when pilot becomes Incapacitated"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"}},"_tags":["story","author_strangattractor","story_46356892"],"author":"strangattractor","children":[46357145],"created_at":"2025-12-22T18:16:48Z","created_at_i":1766427408,"num_comments":1,"objectID":"46356892","points":1,"story_id":46356892,"title":"Emergency Autoland deployed when pilot becomes Incapacitated","updated_at":"2026-03-05T23:16:26Z","url":"https://www.flightradar24.com/blog/aviation-news/aviation-safety/garmin-emergency-autoland-deployed-first-time/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"X4"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"AutoLaTeX: how to automatize LaTeX compilation"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"http://www.arakhne.org/autolatex/"}},"_tags":["story","author_X4","story_6464142"],"author":"X4","created_at":"2013-09-29T04:34:37Z","created_at_i":1380429277,"num_comments":0,"objectID":"6464142","points":1,"story_id":6464142,"title":"AutoLaTeX: how to automatize LaTeX compilation","updated_at":"2024-09-19T20:04:08Z","url":"http://www.arakhne.org/autolatex/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"prostoalex"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Piper gets FAA approval to use Garmin Autoland"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.aopa.org/news-and-media/all-news/2020/may/18/piper-first-with-garmin-autoland-approval"}},"_tags":["story","author_prostoalex","story_23239754"],"author":"prostoalex","created_at":"2020-05-19T20:26:32Z","created_at_i":1589919992,"num_comments":0,"objectID":"23239754","points":1,"story_id":23239754,"title":"Piper gets FAA approval to use Garmin Autoland","updated_at":"2024-09-20T06:13:27Z","url":"https://www.aopa.org/news-and-media/all-news/2020/may/18/piper-first-with-garmin-autoland-approval"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"endymi0n"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Garmin, Cirrus, and Piper Release Emergency Autoland System for Passengers"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.boldmethod.com/learn-to-fly/systems/emergency-autoland-system-for-passengers/"}},"_tags":["story","author_endymi0n","story_21407168"],"author":"endymi0n","created_at":"2019-10-31T10:10:51Z","created_at_i":1572516651,"num_comments":0,"objectID":"21407168","points":1,"story_id":21407168,"title":"Garmin, Cirrus, and Piper Release Emergency Autoland System for Passengers","updated_at":"2024-09-20T05:05:52Z","url":"https://www.boldmethod.com/learn-to-fly/systems/emergency-autoland-system-for-passengers/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jkuria"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"AI meets autoland Germany\u2019s conservative economic model is being put to the test"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.economist.com/news/special-report/21740310-economy-built-caution-must-learn-live-disruption-germanys-conservative"}},"_tags":["story","author_jkuria","story_16840869"],"author":"jkuria","created_at":"2018-04-15T02:56:48Z","created_at_i":1523761008,"num_comments":0,"objectID":"16840869","points":1,"story_id":16840869,"title":"AI meets autoland Germany\u2019s conservative economic model is being put to the test","updated_at":"2024-09-20T02:19:52Z","url":"https://www.economist.com/news/special-report/21740310-economy-built-caution-must-learn-live-disruption-germanys-conservative"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nharada"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hello HN! I want to share something me and a few friends have been working on for a while now \u2014 Zeroshot, a web tool that builds image classifiers using text-image models and autolabeling. What does this mean in practice? You can put together an image classifier in about 30 seconds that\u2019s faster and more accurate than CLIP, but that you can deploy yourself however you\u2019d like. It\u2019s open source, commercially licensed, and doesn\u2019t require you to pay anyone per API call.
Here's a 2 minute video that shows it off: https://www.youtube.com/watch?v=S4R1gtmM-Lo
How/why does it work?\nWe believe that with the rise of foundation vision models, computer vision will fundamentally change. These powerful models will let any devs \u201ccompile\u201d a model ahead of time with a subset of the foundation model\u2019s characteristics, using only text and a web-tool. The days of teams of MLEs building complex models and pipelines are ending.
Zeroshot works by using two powerful pre-trained models, CLIP and DINOv2 together. The web-app allows users to quickly create our training sets via text search. Using pre-cached DINOv2 features, we generate a simple linear model that can be trained and deployed without any fine-tuning. Since you can see what\u2019s going into your training set, you can tune your prompts to get the type of performance or detail you want.
CLIP Small -- Size: 335 MB, Latency: 35ms
CLIP Large -- Size: 891 MB, Latency: 276ms
Zeroshot -- Size: 85 MB, Latency: 20ms
What\u2019s next?\nWe wanna see how people use or would use the tool before deciding what to do next. On the list: clients for iOS and NodeJS, speeding up GPU inference times via TensorRT, offering larger Zeroshot models for better accuracy, easier results refining, support for bringing your own data lake, model refinement using GPT-V, we\u2019ve got plenty of ideas."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Build an open-source computer vision model in seconds using text"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://usezeroshot.com/"}},"_tags":["story","author_nharada","story_38450690","show_hn"],"author":"nharada","children":[38453194,38453544,38455066,38458138,38458535],"created_at":"2023-11-28T20:18:32Z","created_at_i":1701202712,"num_comments":14,"objectID":"38450690","points":64,"story_id":38450690,"story_text":"Hello HN! I want to share something me and a few friends have been working on for a while now \u2014 Zeroshot, a web tool that builds image classifiers using text-image models and autolabeling. What does this mean in practice? You can put together an image classifier in about 30 seconds that\u2019s faster and more accurate than CLIP, but that you can deploy yourself however you\u2019d like. It\u2019s open source, commercially licensed, and doesn\u2019t require you to pay anyone per API call.
Here's a 2 minute video that shows it off: https://www.youtube.com/watch?v=S4R1gtmM-Lo
How/why does it work?\nWe believe that with the rise of foundation vision models, computer vision will fundamentally change. These powerful models will let any devs \u201ccompile\u201d a model ahead of time with a subset of the foundation model\u2019s characteristics, using only text and a web-tool. The days of teams of MLEs building complex models and pipelines are ending.
Zeroshot works by using two powerful pre-trained models, CLIP and DINOv2 together. The web-app allows users to quickly create our training sets via text search. Using pre-cached DINOv2 features, we generate a simple linear model that can be trained and deployed without any fine-tuning. Since you can see what\u2019s going into your training set, you can tune your prompts to get the type of performance or detail you want.
CLIP Small -- Size: 335 MB, Latency: 35ms
CLIP Large -- Size: 891 MB, Latency: 276ms
Zeroshot -- Size: 85 MB, Latency: 20ms
What\u2019s next?\nWe wanna see how people use or would use the tool before deciding what to do next. On the list: clients for iOS and NodeJS, speeding up GPU inference times via TensorRT, offering larger Zeroshot models for better accuracy, easier results refining, support for bringing your own data lake, model refinement using GPT-V, we\u2019ve got plenty of ideas.","title":"Show HN: Build an open-source computer vision model in seconds using text","updated_at":"2024-09-20T15:49:02Z","url":"https://usezeroshot.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"c1yd3i"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"!PHL 01/165 PHL AD AP RDO ALTIMETER UNREL. AUTOLAND, HUD TO TOUCHDOWN, ENHANCED FLT VISION SYSTEMS TO TOUCHDOWN, HEL OPS REQUIRING RDO ALTIMETER DATA TO INCLUDE HOVER AUTOPILOT MODES AND CAT A/B/PERFORMANCE CLASS TKOF AND LDG NOT AUTHORIZED EXC FOR ACFT USING APPROVED ALTERNATIVE METHODS OF COMPLIANCE DUE TO 5G C-BAND INTERFERENCE PLUS SEE AIRWORTHINESS DIRECTIVES 2021-23-12, 2021-23-13 2201190501-2401190501"},"title":{"matchLevel":"none","matchedWords":[],"value":"NOTAM: Radio Altimeter Unreliable Due to 5G C-Band Interference"}},"_tags":["story","author_c1yd3i","story_29938406","ask_hn"],"author":"c1yd3i","children":[29938660],"created_at":"2022-01-14T18:46:19Z","created_at_i":1642185979,"num_comments":2,"objectID":"29938406","points":3,"story_id":29938406,"story_text":"!PHL 01/165 PHL AD AP RDO ALTIMETER UNREL. AUTOLAND, HUD TO TOUCHDOWN, ENHANCED FLT VISION SYSTEMS TO TOUCHDOWN, HEL OPS REQUIRING RDO ALTIMETER DATA TO INCLUDE HOVER AUTOPILOT MODES AND CAT A/B/PERFORMANCE CLASS TKOF AND LDG NOT AUTHORIZED EXC FOR ACFT USING APPROVED ALTERNATIVE METHODS OF COMPLIANCE DUE TO 5G C-BAND INTERFERENCE PLUS SEE AIRWORTHINESS DIRECTIVES 2021-23-12, 2021-23-13 2201190501-2401190501","title":"NOTAM: Radio Altimeter Unreliable Due to 5G C-Band Interference","updated_at":"2024-09-20T10:14:40Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hammock"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"In aviation we already have planes that can and regularly do land themselves without any pilot touching the controls. Landing being the most dangerous part of flying (don\u2019t try to compare it to auto-parallel park) It\u2019s called autoland. The plane\u2019s sensors gather left-right and up-down info and then control the plane down into the runway. We do this with planes carrying as many as 350+ souls on board. You would never even know if a pilot was using autoland, and you\u2019ve probably never heard of an autoland accident the way you have with self-driving.
The difference is, for a plane to do an auto-land (Cat III Landing) you need six things. The airport has to be certified, the crew needs to be certified, the aircraft needs to be certified, the company you are working with must be certified with its home civil aviation authority, the airport/country you are doing the cat III must certify the company as well, AND the weather should actually be low vis (meaning other types of landings like visual, VNAV and standard ILS are more difficult/impossible).
If you have just one of these items not certified, then the pilot can\u2019t do an auto-land.
Until we have the same rigor in self-driving, I would not anticipate much success, for safety reasons."},"title":{"matchLevel":"none","matchedWords":[],"value":"Tell HN: Biggest self-driving challenge is not technology"}},"_tags":["story","author_hammock","story_39790764","ask_hn"],"author":"hammock","children":[39791122,39793361],"created_at":"2024-03-22T13:51:16Z","created_at_i":1711115476,"num_comments":2,"objectID":"39790764","points":2,"story_id":39790764,"story_text":"In aviation we already have planes that can and regularly do land themselves without any pilot touching the controls. Landing being the most dangerous part of flying (don\u2019t try to compare it to auto-parallel park) It\u2019s called autoland. The plane\u2019s sensors gather left-right and up-down info and then control the plane down into the runway. We do this with planes carrying as many as 350+ souls on board. You would never even know if a pilot was using autoland, and you\u2019ve probably never heard of an autoland accident the way you have with self-driving.
The difference is, for a plane to do an auto-land (Cat III Landing) you need six things. The airport has to be certified, the crew needs to be certified, the aircraft needs to be certified, the company you are working with must be certified with its home civil aviation authority, the airport/country you are doing the cat III must certify the company as well, AND the weather should actually be low vis (meaning other types of landings like visual, VNAV and standard ILS are more difficult/impossible).
If you have just one of these items not certified, then the pilot can\u2019t do an auto-land.
Until we have the same rigor in self-driving, I would not anticipate much success, for safety reasons.","title":"Tell HN: Biggest self-driving challenge is not technology","updated_at":"2024-09-20T16:42:26Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ramacastro"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hey everyone!
I created a startup with a friend to make it easier for anyone to create AIs without any prior experience or knowledge in the field. We started with computer vision, specifically object detection, but we plan to add all subfields of AI eventually, like natural language processing, speech and all the cool generative stuff.
You can do everything inside the platform: create datasets, train artificial neural networks and use them to give intelligence to your web apps, mobile apps, machines, security cameras, robots, drones, cars and everything else.
It is free to use forever without any credit card required. You can connect google colab notebooks to our platform and use their free GPUs for training. Forget about all the different labeling file formats and running complex code yourself, just connect colab to Theos, click the Start Training button and the metrics will stream directly to your browser, and we will automatically upload weights checkpoints every epoch so you don't have to worry about colab shutting down and losing all your AI training progress. Also, we have an autolabeler feature to let your current AI's version help you make a better version of itself 100 times faster than labeling manually. Finally, you can deploy your AI as a rest API and use it in your software with simple http requests.
Here's the getting started guide for object detection:
https://docs.theos.ai/get-started/object-detection
Please let me know if you find it helpful and what subfield of AI you think we should tackle next.
(the landing page is not finished yet, soon we will release a way better version, but for now you can sign up and use the app for free)
Thanks!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Theos AI - development platform for AI"}},"_tags":["story","author_ramacastro","story_32038181","show_hn"],"author":"ramacastro","created_at":"2022-07-09T18:56:03Z","created_at_i":1657392963,"num_comments":0,"objectID":"32038181","points":1,"story_id":32038181,"story_text":"Hey everyone!
I created a startup with a friend to make it easier for anyone to create AIs without any prior experience or knowledge in the field. We started with computer vision, specifically object detection, but we plan to add all subfields of AI eventually, like natural language processing, speech and all the cool generative stuff.
You can do everything inside the platform: create datasets, train artificial neural networks and use them to give intelligence to your web apps, mobile apps, machines, security cameras, robots, drones, cars and everything else.
It is free to use forever without any credit card required. You can connect google colab notebooks to our platform and use their free GPUs for training. Forget about all the different labeling file formats and running complex code yourself, just connect colab to Theos, click the Start Training button and the metrics will stream directly to your browser, and we will automatically upload weights checkpoints every epoch so you don't have to worry about colab shutting down and losing all your AI training progress. Also, we have an autolabeler feature to let your current AI's version help you make a better version of itself 100 times faster than labeling manually. Finally, you can deploy your AI as a rest API and use it in your software with simple http requests.
Here's the getting started guide for object detection:
https://docs.theos.ai/get-started/object-detection
Please let me know if you find it helpful and what subfield of AI you think we should tackle next.
(the landing page is not finished yet, soon we will release a way better version, but for now you can sign up and use the app for free)
Thanks!","title":"Show HN: Theos AI - development platform for AI","updated_at":"2024-09-20T11:30:57Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"wilonth"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: An \u201cautomatic\u201d way to learn Chinese"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://play.google.com/store/apps/details?id=gupik.autolang.zh"}},"_tags":["story","author_wilonth","story_18463886","show_hn"],"author":"wilonth","children":[18464421,18465365,18465683,18465824,18465985,18466333,18467409,18467513,18467894,18468491,18468755,18468864,18469700,18469796,18470070,18472357,18473550,18473795,18476076],"created_at":"2018-11-15T22:04:32Z","created_at_i":1542319472,"num_comments":38,"objectID":"18463886","points":57,"story_id":18463886,"title":"Show HN: An \u201cautomatic\u201d way to learn Chinese","updated_at":"2024-09-21T17:04:19Z","url":"https://play.google.com/store/apps/details?id=gupik.autolang.zh"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Nycto"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"matchLevel":"none","matchedWords":[],"value":"A radically new type of engine"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"http://www.autoblog.com/2009/11/30/autoline-on-autoblog-with-john-mcelroy/"}},"_tags":["story","author_Nycto","story_972846"],"author":"Nycto","children":[973089,973178,973248,973729,973806,973925,974088],"created_at":"2009-12-02T20:28:19Z","created_at_i":1259785699,"num_comments":14,"objectID":"972846","points":56,"story_id":972846,"story_text":"","title":"A radically new type of engine","updated_at":"2023-09-06T20:17:05Z","url":"http://www.autoblog.com/2009/11/30/autoline-on-autoblog-with-john-mcelroy/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Shivetya"},"title":{"matchLevel":"none","matchedWords":[],"value":"The race for autonomous cars is over. Silicon Valley lost"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"http://www.autoblog.com/2017/02/21/race-for-autonomous-cars-is-over-mcelroy-autoline-opinion/"}},"_tags":["story","author_Shivetya","story_13704836"],"author":"Shivetya","children":[13705010,13705050,13705067,13705077,13705273,13705275,13705580,13705609,13705931,13706149],"created_at":"2017-02-22T11:58:31Z","created_at_i":1487764711,"num_comments":17,"objectID":"13704836","points":29,"story_id":13704836,"title":"The race for autonomous cars is over. Silicon Valley lost","updated_at":"2024-09-20T00:22:24Z","url":"http://www.autoblog.com/2017/02/21/race-for-autonomous-cars-is-over-mcelroy-autoline-opinion/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kaudinya"},"title":{"matchLevel":"none","matchedWords":[],"value":"dstack.ai \u2013 Collaborate on R-generated results without web development skills"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/collaborate-on-r-generated-results-without-web-development-skills-62a510316f14?source=email-4be6ceea0140-1587062511402-layerCake.autoLayerCakeWriterNotification-------------------------92ffc4d7_0770_42c0_bc80_25d79bdde974&sk=c77cd29cf104d72788e6b0868c3d9425."}},"_tags":["story","author_kaudinya","story_22899337"],"author":"kaudinya","created_at":"2020-04-17T14:41:18Z","created_at_i":1587134478,"num_comments":0,"objectID":"22899337","points":6,"story_id":22899337,"title":"dstack.ai \u2013 Collaborate on R-generated results without web development skills","updated_at":"2024-09-20T06:06:55Z","url":"https://towardsdatascience.com/collaborate-on-r-generated-results-without-web-development-skills-62a510316f14?source=email-4be6ceea0140-1587062511402-layerCake.autoLayerCakeWriterNotification-------------------------92ffc4d7_0770_42c0_bc80_25d79bdde974&sk=c77cd29cf104d72788e6b0868c3d9425."},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"emudobs"},"title":{"matchLevel":"none","matchedWords":[],"value":"Building a Matplotlib to Plotly Chart Conversion Tool"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/matplotlib-to-plotly-chart-conversion-4bd260e73434?source=email-d25c490901f-1589156525839-layerCake.autoLayerCakeWriterNotification-------------------------ec4b69f6_d8b9_4994_b136_451e5d146a6c&sk=4e38e17d877b9dbf25fa65c28bbd8753"}},"_tags":["story","author_emudobs","story_23142306"],"author":"emudobs","created_at":"2020-05-11T14:10:26Z","created_at_i":1589206226,"num_comments":0,"objectID":"23142306","points":4,"story_id":23142306,"title":"Building a Matplotlib to Plotly Chart Conversion Tool","updated_at":"2024-09-20T06:11:58Z","url":"https://towardsdatascience.com/matplotlib-to-plotly-chart-conversion-4bd260e73434?source=email-d25c490901f-1589156525839-layerCake.autoLayerCakeWriterNotification-------------------------ec4b69f6_d8b9_4994_b136_451e5d146a6c&sk=4e38e17d877b9dbf25fa65c28bbd8753"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"zwilliamson"},"title":{"matchLevel":"none","matchedWords":[],"value":"Automating AWS CloudWatch Logs on Ubuntu"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autozane.com/automating-aws-cloudwatch-logs-on-ubuntu/"}},"_tags":["story","author_zwilliamson","story_12537044"],"author":"zwilliamson","created_at":"2016-09-20T04:23:44Z","created_at_i":1474345424,"num_comments":0,"objectID":"12537044","points":4,"story_id":12537044,"title":"Automating AWS CloudWatch Logs on Ubuntu","updated_at":"2024-09-19T23:42:55Z","url":"https://autozane.com/automating-aws-cloudwatch-logs-on-ubuntu/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"altilunium"},"title":{"matchLevel":"none","matchedWords":[],"value":"Plane makes safe emergency landing in Colorado without a pilot's help"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://www.cbsnews.com/colorado/news/plane-emergency-landing-colorado-autoland/"}},"_tags":["story","author_altilunium","story_46376331"],"author":"altilunium","created_at":"2025-12-24T15:20:45Z","created_at_i":1766589645,"num_comments":0,"objectID":"46376331","points":3,"story_id":46376331,"title":"Plane makes safe emergency landing in Colorado without a pilot's help","updated_at":"2026-03-05T23:17:49Z","url":"https://www.cbsnews.com/colorado/news/plane-emergency-landing-colorado-autoland/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"HIP_HOP"},"title":{"matchLevel":"none","matchedWords":[],"value":"Pandas Analytics Server"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/pandas-analytics-server-d9abceec888b?source=email-68bf0657a0cd-1583181295657-layerCake.autoLayerCakeWriterNotification-------------------------1f574b34_38e1_44e9_8e80_b99df8c01286&sk=937d40a6d1be68cc3b2a852475f4bb40"}},"_tags":["story","author_HIP_HOP","story_22468726"],"author":"HIP_HOP","created_at":"2020-03-02T20:43:48Z","created_at_i":1583181828,"num_comments":0,"objectID":"22468726","points":3,"story_id":22468726,"title":"Pandas Analytics Server","updated_at":"2024-09-20T05:44:27Z","url":"https://towardsdatascience.com/pandas-analytics-server-d9abceec888b?source=email-68bf0657a0cd-1583181295657-layerCake.autoLayerCakeWriterNotification-------------------------1f574b34_38e1_44e9_8e80_b99df8c01286&sk=937d40a6d1be68cc3b2a852475f4bb40"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hoansdz"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"Do you think it's worth pursuing ?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: A lightweight compiler for untrusted AI Agent scripts"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autolang.vercel.app/docs/philosophy-vision"}},"_tags":["story","author_hoansdz","story_48336380","show_hn"],"author":"hoansdz","children":[48336685],"created_at":"2026-05-30T14:09:27Z","created_at_i":1780150167,"num_comments":2,"objectID":"48336380","points":2,"story_id":48336380,"story_text":"Do you think it's worth pursuing ?","title":"Show HN: A lightweight compiler for untrusted AI Agent scripts","updated_at":"2026-05-31T15:30:30Z","url":"https://autolang.vercel.app/docs/philosophy-vision"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hoansdz"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: I made a compiler/VM for untrusted scripts"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autolang.vercel.app/docs/philosophy-vision"}},"_tags":["story","author_hoansdz","story_48266073","show_hn"],"author":"hoansdz","children":[48266087],"created_at":"2026-05-25T12:31:29Z","created_at_i":1779712289,"num_comments":0,"objectID":"48266073","points":2,"story_id":48266073,"title":"Show HN: I made a compiler/VM for untrusted scripts","updated_at":"2026-05-25T12:42:05Z","url":"https://autolang.vercel.app/docs/philosophy-vision"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pinter69"},"title":{"matchLevel":"none","matchedWords":[],"value":"Machine Learning project management \u2013 A decision makers\u2019 guide"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/machine-learning-project-management-a-decision-makers-guide-2d3d-ai-6fbd59504bb1?source=email-b88fb602cc96-1586788460285-layerCake.autoLayerCakeWriterNotification-------------------------15bc1e31_5c04_4e39_a286_12b192b33dc5&sk=a37c1564937fdd487840b696e58d406f"}},"_tags":["story","author_pinter69","story_22863051"],"author":"pinter69","created_at":"2020-04-14T05:41:25Z","created_at_i":1586842885,"num_comments":0,"objectID":"22863051","points":2,"story_id":22863051,"title":"Machine Learning project management \u2013 A decision makers\u2019 guide","updated_at":"2024-09-20T06:03:28Z","url":"https://towardsdatascience.com/machine-learning-project-management-a-decision-makers-guide-2d3d-ai-6fbd59504bb1?source=email-b88fb602cc96-1586788460285-layerCake.autoLayerCakeWriterNotification-------------------------15bc1e31_5c04_4e39_a286_12b192b33dc5&sk=a37c1564937fdd487840b696e58d406f"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"HIP_HOP"},"title":{"matchLevel":"none","matchedWords":[],"value":"Don\u2019t repeat my mistakes while developing a Machine Learning library"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/dont-repeat-my-mistakes-while-developing-a-machine-learning-library-21b79975d23?source=email-68bf0657a0cd-1585680057567-layerCake.autoLayerCakeWriterNotification-------------------------a70a9897_7385_4a38_9c05_5557a2794a20&sk=49733fad4ceadea4d86a0bae32cda1e1"}},"_tags":["story","author_HIP_HOP","story_22740956"],"author":"HIP_HOP","created_at":"2020-03-31T18:42:59Z","created_at_i":1585680179,"num_comments":0,"objectID":"22740956","points":2,"story_id":22740956,"title":"Don\u2019t repeat my mistakes while developing a Machine Learning library","updated_at":"2024-09-20T05:52:41Z","url":"https://towardsdatascience.com/dont-repeat-my-mistakes-while-developing-a-machine-learning-library-21b79975d23?source=email-68bf0657a0cd-1585680057567-layerCake.autoLayerCakeWriterNotification-------------------------a70a9897_7385_4a38_9c05_5557a2794a20&sk=49733fad4ceadea4d86a0bae32cda1e1"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"zwilliamson"},"title":{"matchLevel":"none","matchedWords":[],"value":"Managing Terraform Versions on Jenkins"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://autozane.com/managing-terraform-versions-on-jenkins/"}},"_tags":["story","author_zwilliamson","story_12485638"],"author":"zwilliamson","created_at":"2016-09-13T04:29:51Z","created_at_i":1473740991,"num_comments":0,"objectID":"12485638","points":2,"story_id":12485638,"title":"Managing Terraform Versions on Jenkins","updated_at":"2024-09-19T23:44:37Z","url":"https://autozane.com/managing-terraform-versions-on-jenkins/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rcshubhadeep"},"title":{"matchLevel":"none","matchedWords":[],"value":"Surprise and the Tale of Entropy"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/surprise-and-the-tale-of-entropy-c127da84937a?source=email-75704da7f0ae-1589298944905-layerCake.autoLayerCakeWriterNotification-------------------------fb05ada9_68a3_4fa4_a0a1_c221b3c8614e&sk=6ad8234529a459fbd8c288eea7f433c3."}},"_tags":["story","author_rcshubhadeep","story_23165726"],"author":"rcshubhadeep","created_at":"2020-05-13T11:22:57Z","created_at_i":1589368977,"num_comments":0,"objectID":"23165726","points":1,"story_id":23165726,"title":"Surprise and the Tale of Entropy","updated_at":"2024-09-20T06:14:08Z","url":"https://towardsdatascience.com/surprise-and-the-tale-of-entropy-c127da84937a?source=email-75704da7f0ae-1589298944905-layerCake.autoLayerCakeWriterNotification-------------------------fb05ada9_68a3_4fa4_a0a1_c221b3c8614e&sk=6ad8234529a459fbd8c288eea7f433c3."},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bweber"},"title":{"matchLevel":"none","matchedWords":[],"value":"DevOps for Data Science with GKE and Stackdriver"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/devops-for-data-science-with-gcp-3e6b5c3dd4f6?source=email-a80e1f69e782-1581980059142-layerCake.autoLayerCakeWriterNotification-------------------------77b835c2_738e_404a_a704_e8a415fcf7e0&sk=a4bd2710af0b1dfca4c5fa9dfd21f913"}},"_tags":["story","author_bweber","story_22356730"],"author":"bweber","created_at":"2020-02-18T15:31:55Z","created_at_i":1582039915,"num_comments":0,"objectID":"22356730","points":1,"story_id":22356730,"title":"DevOps for Data Science with GKE and Stackdriver","updated_at":"2024-09-20T05:42:19Z","url":"https://towardsdatascience.com/devops-for-data-science-with-gcp-3e6b5c3dd4f6?source=email-a80e1f69e782-1581980059142-layerCake.autoLayerCakeWriterNotification-------------------------77b835c2_738e_404a_a704_e8a415fcf7e0&sk=a4bd2710af0b1dfca4c5fa9dfd21f913"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"HIP_HOP"},"title":{"matchLevel":"none","matchedWords":[],"value":"New Features in Pandas 1.0"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"https://towardsdatascience.com/5-new-features-in-pandas-1-0-you-should-know-about-fc31c83e396b?source=email-68bf0657a0cd-1580749467218-layerCake.autoLayerCakeWriterNotification-------------------------576a67b4_fcbd_4392_8ab6_11137f8f98b5&sk=e20b65134a9a908de33336e3889b09bd"}},"_tags":["story","author_HIP_HOP","story_22244906"],"author":"HIP_HOP","created_at":"2020-02-05T08:12:53Z","created_at_i":1580890373,"num_comments":0,"objectID":"22244906","points":1,"story_id":22244906,"title":"New Features in Pandas 1.0","updated_at":"2024-09-20T05:40:05Z","url":"https://towardsdatascience.com/5-new-features-in-pandas-1-0-you-should-know-about-fc31c83e396b?source=email-68bf0657a0cd-1580749467218-layerCake.autoLayerCakeWriterNotification-------------------------576a67b4_fcbd_4392_8ab6_11137f8f98b5&sk=e20b65134a9a908de33336e3889b09bd"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"chaostheory"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ulteo Virtual Desktop: your Linux applications on Windows"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"http://www.ulteo.com/home/en/virtualdesktop?autolang=en"}},"_tags":["story","author_chaostheory","story_194787"],"author":"chaostheory","created_at":"2008-05-20T06:35:17Z","created_at_i":1211265317,"num_comments":0,"objectID":"194787","points":1,"story_id":194787,"title":"Ulteo Virtual Desktop: your Linux applications on Windows","updated_at":"2023-09-06T19:15:10Z","url":"http://www.ulteo.com/home/en/virtualdesktop?autolang=en"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pseudolus"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Boeing flags 737 MAX software glitch affecting some automated approach functions"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.cnbc.com/2026/09/26/boeing-737-max-navigation-software-glitch.html"}},"_tags":["story","author_pseudolus","story_49869809"],"author":"pseudolus","created_at":"2026-09-27T19:08:31Z","created_at_i":1790536111,"num_comments":0,"objectID":"49869809","points":6,"story_id":49869809,"title":"Boeing flags 737 MAX software glitch affecting some automated approach functions","updated_at":"2026-09-28T00:37:17Z","url":"https://www.cnbc.com/2026/09/26/boeing-737-max-navigation-software-glitch.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sidharthkmenon"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hello! We\u2019re Sid, Alex, Ketan, and Milan. We\u2019re building Whiteboard (https://whiteboard.dev.fast/), an open-source desktop app where humans and agents can architect software together in a common workspace. Here\u2019s our repo: https://github.com/devdotfast/whiteboard.
We were missing the feeling of a \u201cwhiteboard session\u201d with another dev where you leave with a deep understanding of a system, so we built this app for ourselves. Whiteboard plugs into the tools you already use - e.g. Claude Code, Codex, etc. \u2013 and gives your agent an SDK to draw on an in-app canvas to describe its work. We began with an MVP based on HTML artifacts and started rethinking the app as we ran into limitations:
1. Built on top of CodeOSS: We found that in pure HTML tools it was hard to connect a spec or diagram to code. In Whiteboard, when you click on visualizations like a sequence diagram, an entity relationship diagram, or a quote from the agent\u2019s trace, you can jump to the underlying code directly. When navigating code, you get keybindings and LSP support from VSCode out of the box. We\u2019ve found this is especially valuable because tradeoffs are often only discovered after a first pass at implementation (re: slop)
2. Semantic diff viewer: we wrote a semantic, AST-aware diff viewer in Rust so you can only view the code changes which are relevant to you [1]. We\u2019ve set up some sane defaults: large added functions are summarized as pseudocode, and things like unit tests and large documentation changes are collapsed / hidden. This is all customizable with a WASM-based plugin system.
3. Decision Log: We found it difficult to reason about what set of decisions our agents made autonomously. So we built tools for agents to query and link their own traces to the Whiteboard, so you can understand how the requirements that you set were implemented, and understand what decisions your agent made autonomously.
Here\u2019s a quick demo video explaining more: https://www.youtube.com/watch?v=ChPn3ftULWE
Folks at companies like Salesforce and Modal are using Whiteboard today as a review tool for architecture or spec-level changes \u2013 really any change where they want to be involved:
1. Reviewing your own coding agent\u2019s work: because Whiteboard makes it easier to review large amounts of code, folks will typically have their AI agents create a prototype and a corresponding Whiteboard session so they can iterate on the design.
2. Reviewing other people\u2019s changes: We\u2019ve found that Whiteboard is particularly helpful when composed with tools like Greptile. For example, you can run an automated code reviewer on small changes and escalate to a Whiteboard session for the changes that require human judgement.
Why we built this: we\u2019re four buddies from college who quit our jobs as tech leads right before agentic coding became industry standard. As we iterated towards an MVP for a previous idea, we struggled to maintain a comprehensible codebase while reaping all the velocity benefits of agentic coding. As more PRs were merged without our understanding, we felt a \u2018cognitive debt\u2019 begin to seep in, until it became difficult for us to even contribute to the system [2].
We\u2019re releasing our desktop app under an MIT license. Please poke through and feel free to contribute! Eventually we\u2019ll charge companies for a hosted web version that manages whiteboard session creation alongside features like trajectory storage and multiplayer reviews. Everything will always remain self-hostable.
Thanks for reading, and we hope you try it out! We would love to hear any feedback and to learn from your expertise.
Here\u2019s are the project links again: https://github.com/devdotfast/whiteboard, and you can install (for MacOS + Linux) at https://install.dev.fast
[1] diffs library: https://github.com/devdotfast/diffr\n[2] Credit for the term \u2018cognitive debt\u2019 goes to https://www.geoffreylitt.com/2026/07/02/understanding-is-the..."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Whiteboard (YC W26) \u2013 An open-source IDE for thoughtful software design"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/devdotfast/whiteboard"}},"_tags":["story","author_sidharthkmenon","story_49833867","show_hn"],"author":"sidharthkmenon","children":[49834369,49834380,49834508,49834543,49834622,49834714,49834864,49834891,49834901,49834994,49835068,49835228,49835239,49835319,49835627,49835681,49835732,49835940,49835980,49836158,49836271,49836332,49836385,49836389,49836440,49836484,49836507,49836542,49836566,49836767,49836793,49836840,49837428,49837464,49837529,49837645,49837784,49837945,49838273,49838730,49838809,49838943,49839050,49840151,49840172,49840326,49840367,49840504,49840531,49840608,49840667,49840763,49840900,49841065,49841081,49841463,49841490,49841932,49842013,49842353,49842572,49842763,49843227,49846422,49849240,49851442,49852471,49854301,49858622,49859426,49860478,49863075],"created_at":"2026-09-24T17:21:36Z","created_at_i":1790270496,"num_comments":141,"objectID":"49833867","points":419,"story_id":49833867,"story_text":"Hello! We\u2019re Sid, Alex, Ketan, and Milan. We\u2019re building Whiteboard (https://whiteboard.dev.fast/), an open-source desktop app where humans and agents can architect software together in a common workspace. Here\u2019s our repo: https://github.com/devdotfast/whiteboard.
We were missing the feeling of a \u201cwhiteboard session\u201d with another dev where you leave with a deep understanding of a system, so we built this app for ourselves. Whiteboard plugs into the tools you already use - e.g. Claude Code, Codex, etc. \u2013 and gives your agent an SDK to draw on an in-app canvas to describe its work. We began with an MVP based on HTML artifacts and started rethinking the app as we ran into limitations:
1. Built on top of CodeOSS: We found that in pure HTML tools it was hard to connect a spec or diagram to code. In Whiteboard, when you click on visualizations like a sequence diagram, an entity relationship diagram, or a quote from the agent\u2019s trace, you can jump to the underlying code directly. When navigating code, you get keybindings and LSP support from VSCode out of the box. We\u2019ve found this is especially valuable because tradeoffs are often only discovered after a first pass at implementation (re: slop)
2. Semantic diff viewer: we wrote a semantic, AST-aware diff viewer in Rust so you can only view the code changes which are relevant to you [1]. We\u2019ve set up some sane defaults: large added functions are summarized as pseudocode, and things like unit tests and large documentation changes are collapsed / hidden. This is all customizable with a WASM-based plugin system.
3. Decision Log: We found it difficult to reason about what set of decisions our agents made autonomously. So we built tools for agents to query and link their own traces to the Whiteboard, so you can understand how the requirements that you set were implemented, and understand what decisions your agent made autonomously.
Here\u2019s a quick demo video explaining more: https://www.youtube.com/watch?v=ChPn3ftULWE
Folks at companies like Salesforce and Modal are using Whiteboard today as a review tool for architecture or spec-level changes \u2013 really any change where they want to be involved:
1. Reviewing your own coding agent\u2019s work: because Whiteboard makes it easier to review large amounts of code, folks will typically have their AI agents create a prototype and a corresponding Whiteboard session so they can iterate on the design.
2. Reviewing other people\u2019s changes: We\u2019ve found that Whiteboard is particularly helpful when composed with tools like Greptile. For example, you can run an automated code reviewer on small changes and escalate to a Whiteboard session for the changes that require human judgement.
Why we built this: we\u2019re four buddies from college who quit our jobs as tech leads right before agentic coding became industry standard. As we iterated towards an MVP for a previous idea, we struggled to maintain a comprehensible codebase while reaping all the velocity benefits of agentic coding. As more PRs were merged without our understanding, we felt a \u2018cognitive debt\u2019 begin to seep in, until it became difficult for us to even contribute to the system [2].
We\u2019re releasing our desktop app under an MIT license. Please poke through and feel free to contribute! Eventually we\u2019ll charge companies for a hosted web version that manages whiteboard session creation alongside features like trajectory storage and multiplayer reviews. Everything will always remain self-hostable.
Thanks for reading, and we hope you try it out! We would love to hear any feedback and to learn from your expertise.
Here\u2019s are the project links again: https://github.com/devdotfast/whiteboard, and you can install (for MacOS + Linux) at https://install.dev.fast
[1] diffs library: https://github.com/devdotfast/diffr\n[2] Credit for the term \u2018cognitive debt\u2019 goes to https://www.geoffreylitt.com/2026/07/02/understanding-is-the...","title":"Show HN: Whiteboard (YC W26) \u2013 An open-source IDE for thoughtful software design","updated_at":"2026-09-28T07:09:34Z","url":"https://github.com/devdotfast/whiteboard"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ZaharaHussain"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Hi HN, I built ai\u00b7rete\u00b7rag because I kept seeing teams put an LLM in charge of decisions that need to be auditable (lending, fraud, clinical triage), then bolt on "guardrails" after the fact.
It runs the two in series instead:
1. A pure-Python Rete engine evaluates YAML rules against your facts. The verdict comes only from here. Same facts, same verdict, every time, with salience-based conflict resolution.\n2. RAG retrieves passages from your own policy documents, and an LLM writes a plain-English explanation of the decision that was already made, citing those passages. It can't change the verdict.
A few things that went further than I expected:\n- Rules are a graph, not flat lists: nested all/any/not, and rules can assert facts that other rules consume (forward chaining). The decision trace shows the causal chain.\n- Audit mode records every rule evaluated, including the ones that didn't fire, condition by condition, with a snapshot of the rule set for replay.\n- Rules can steer retrieval (a fired rule narrows which documents get searched), and retrieved text can be turned into facts for the engine.\n- Non-technical authors can build rules in a visual editor, or paste a policy document and get LLM-drafted rules with citations. Drafts are never saved without review. YAML is still there for engineers.
The landing page has a live demo with no signup (8 demo domains: loan, fraud, clinical, insurance, legal, ops, e-commerce, blockchain). There's also an MCP server, so Claude and other agents can call /decide as a tool: `uvx ai-rete-rag-mcp`.
To be upfront: it's a hosted product with a free tier. The MCP client is open source (MIT, github.com/zaharajabeen13-create/ai-rete-rag-mcp); the engine and platform are not open source right now.
I'd especially like to hear from anyone who has had to explain an automated decision to a regulator or an auditor: what did they actually ask for?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: AI\u00b7rete\u00b7RAG \u2013 a Rete rule engine decides, RAG explains why"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://ai-rete-rag.com/"}},"_tags":["story","author_ZaharaHussain","story_49803683","show_hn"],"author":"ZaharaHussain","children":[49805272,49806089,49806185,49806638,49811819,49812349,49814308,49816208,49818067,49824723,49826922,49852224],"created_at":"2026-09-22T16:15:06Z","created_at_i":1790093706,"num_comments":9,"objectID":"49803683","points":44,"story_id":49803683,"story_text":"Hi HN, I built ai\u00b7rete\u00b7rag because I kept seeing teams put an LLM in charge of decisions that need to be auditable (lending, fraud, clinical triage), then bolt on "guardrails" after the fact.
It runs the two in series instead:
1. A pure-Python Rete engine evaluates YAML rules against your facts. The verdict comes only from here. Same facts, same verdict, every time, with salience-based conflict resolution.\n2. RAG retrieves passages from your own policy documents, and an LLM writes a plain-English explanation of the decision that was already made, citing those passages. It can't change the verdict.
A few things that went further than I expected:\n- Rules are a graph, not flat lists: nested all/any/not, and rules can assert facts that other rules consume (forward chaining). The decision trace shows the causal chain.\n- Audit mode records every rule evaluated, including the ones that didn't fire, condition by condition, with a snapshot of the rule set for replay.\n- Rules can steer retrieval (a fired rule narrows which documents get searched), and retrieved text can be turned into facts for the engine.\n- Non-technical authors can build rules in a visual editor, or paste a policy document and get LLM-drafted rules with citations. Drafts are never saved without review. YAML is still there for engineers.
The landing page has a live demo with no signup (8 demo domains: loan, fraud, clinical, insurance, legal, ops, e-commerce, blockchain). There's also an MCP server, so Claude and other agents can call /decide as a tool: `uvx ai-rete-rag-mcp`.
To be upfront: it's a hosted product with a free tier. The MCP client is open source (MIT, github.com/zaharajabeen13-create/ai-rete-rag-mcp); the engine and platform are not open source right now.
I'd especially like to hear from anyone who has had to explain an automated decision to a regulator or an auditor: what did they actually ask for?","title":"Show HN: AI\u00b7rete\u00b7RAG \u2013 a Rete rule engine decides, RAG explains why","updated_at":"2026-09-27T22:49:03Z","url":"https://ai-rete-rag.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"platevoltage"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"Just to preface, this isn't the typical "Am I being replaced by vibe coding" post.
I just had a talk with one of my first ongoing clients, who sort of helped me kickstart my freelancing "career". He is one of those people who has a broad understanding of the tech out there, and what its capabilities are, but is not a programmer. He wanted to show me, and get my opinion on a restaurant management app that he's been using Lovable to develop. Admittedly, a lot of it is very impressive, and he has obviously become very good at using these tools. Part of that scares me, but I've largely accepted the reality that my skillset has to expand beyond being a React code monkey (wish is not what I am, but you get the point). I've always been a curious generalist, so picking up new stuff comes easily, and I'll manage.
Anyways, here's the part that feels really gross. He's working on a feature that does some automated marketing using Higgsfield, so automated slop posting. Also a feature that consumes reviews from vendors, and he intends to use agents to post reviews as well.
I'm 43, I managed to convince my parents to get AOL when I was 12. I made a Geocities site at 14. I've been obsessed with computers and the internet for my entire life. I'm also an artist who had a somewhat successful run as a touring musician, much to the detriment of my career growth (not much money in Punk).
I feel the hobbies that I've been nurturing for my entire life are being destroyed by people who have finally found an entry point that requires the smallest amount of effort. We are literally watching the dead internet happen in real time, social media turning into a AI slop gallery, AI music topping the charts, and I just had a one-on-one conversation with someone that just doesn't care, because they get to feel creative for possibly the first time ever.
How are you all dealing with this?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: I just talked to an AI-obsessed client, and I need a shower afterwards"}},"_tags":["story","author_platevoltage","story_49826029","ask_hn"],"author":"platevoltage","children":[49826104,49826345,49826586,49827123,49827430,49827618,49828265,49828397,49829428,49830222,49830260,49831601,49832296,49832929,49833187,49833692,49833724,49834071,49835479,49835679,49839341,49840334,49841025,49843960,49844439,49845304,49845687,49846939],"created_at":"2026-09-24T03:48:41Z","created_at_i":1790221721,"num_comments":49,"objectID":"49826029","points":30,"story_id":49826029,"story_text":"Just to preface, this isn't the typical "Am I being replaced by vibe coding" post.
I just had a talk with one of my first ongoing clients, who sort of helped me kickstart my freelancing "career". He is one of those people who has a broad understanding of the tech out there, and what its capabilities are, but is not a programmer. He wanted to show me, and get my opinion on a restaurant management app that he's been using Lovable to develop. Admittedly, a lot of it is very impressive, and he has obviously become very good at using these tools. Part of that scares me, but I've largely accepted the reality that my skillset has to expand beyond being a React code monkey (wish is not what I am, but you get the point). I've always been a curious generalist, so picking up new stuff comes easily, and I'll manage.
Anyways, here's the part that feels really gross. He's working on a feature that does some automated marketing using Higgsfield, so automated slop posting. Also a feature that consumes reviews from vendors, and he intends to use agents to post reviews as well.
I'm 43, I managed to convince my parents to get AOL when I was 12. I made a Geocities site at 14. I've been obsessed with computers and the internet for my entire life. I'm also an artist who had a somewhat successful run as a touring musician, much to the detriment of my career growth (not much money in Punk).
I feel the hobbies that I've been nurturing for my entire life are being destroyed by people who have finally found an entry point that requires the smallest amount of effort. We are literally watching the dead internet happen in real time, social media turning into a AI slop gallery, AI music topping the charts, and I just had a one-on-one conversation with someone that just doesn't care, because they get to feel creative for possibly the first time ever.
How are you all dealing with this?","title":"Ask HN: I just talked to an AI-obsessed client, and I need a shower afterwards","updated_at":"2026-09-28T02:28:18Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Alexadar"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autolane"],"value":"I\u2019m a senior ML/Dev with top accounts on freelance platforms. I cannot land a gig because automated platforms have flooded clients\u2019 attention, despite being able to for years. Clients complain as well that they cannot find a workforce because of the AI slop shield. Imagine, you post a job on Upwork and bots get you on LinkedIn in a few minutes. What do you guys do? How do you search for gigs now?"},"title":{"matchLevel":"none","matchedWords":[],"value":"How to Freelance Now?"}},"_tags":["story","author_Alexadar","story_49856000","ask_hn"],"author":"Alexadar","children":[49856254,49856421,49856456,49857452,49858228,49859170,49860525,49860703,49863499,49864196,49865523,49869822],"created_at":"2026-09-26T12:46:34Z","created_at_i":1790426794,"num_comments":15,"objectID":"49856000","points":18,"story_id":49856000,"story_text":"I\u2019m a senior ML/Dev with top accounts on freelance platforms. I cannot land a gig because automated platforms have flooded clients\u2019 attention, despite being able to for years. Clients complain as well that they cannot find a workforce because of the AI slop shield. Imagine, you post a job on Upwork and bots get you on LinkedIn in a few minutes. What do you guys do? How do you search for gigs now?","title":"How to Freelance Now?","updated_at":"2026-09-28T07:01:07Z"}],"hitsPerPage":50,"nbHits":52,"nbPages":2,"page":0,"params":"query=Autolane&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":7,"processingTimingsMS":{"_request":{"roundTrip":23},"afterFetch":{"format":{"highlighting":2,"total":2},"merge":{"mergeLoop":{"prepareNextHit":2,"total":2},"total":2},"total":2},"fetch":{"query":3,"total":4},"total":7},"query":"Autolane","serverTimeMS":10}