[ { "id": 202, "name": "PagerDuty", "slug": "pagerduty", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/18fd6f870541d2398827dedc9e57678e8770c424.png", "website": "http://pagerduty.com", "all_locations": "San Francisco, CA, USA", "long_description": "PagerDuty is an operations performance platform delivering visibility and actionable intelligence across the entire incident lifecycle. Their SaaS-based solution empowers over 10,000 small, mid-size and enterprise global customers such as Comcast, eHarmony, Slack and Lululemon with the insight to intelligently respond to critical disruptions for exceptional customer experience. PagerDuty was founded to deliver a new and innovative approach to increase business response and efficiency. When brand reputation depends on customer satisfaction, PagerDuty arms businesses with the insight to proactively manage incidents and events that may impact customers across their IT environment.", "one_liner": "Real-time visibility into critical apps and services all in one place.", "team_size": 950, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1326790252, "tags": [ "DevOps", "Monitoring" ], "tags_highlighted": [], "top_company": true, "isHiring": false, "nonprofit": false, "batch": "Summer 2010", "status": "Public", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Growth", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/pagerduty", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2010/pagerduty.json", "year_founded": null, "city": "San Francisco", "country": "US", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/pagerduty", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/18fd6f870541d2398827dedc9e57678e8770c424.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 5098, "full_name": "Alex Solomon", "title": "Founder/CTO", "founder_bio": "Co-Founder, PagerDuty. Alex started PagerDuty in 2009 as founding CEO, leading the company through several stages of growth, including product-market fit, fundraising, and expansion of the product vision. He replaced himself in 2016 and stepped into the role of CTO (until 2020), and later, GM of Platform (until 2024). He has served on the PagerDuty board of directors since 2010.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/alexsolo", "x_url": null, "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4153b6619f8644f8eec9f49eb9b1e68299a466e6.jpg" } ], "jobs": [], "news": [ { "title": "Toronto-founded PagerDuty goes public, sees stocks jump in market debut | BetaKit", "url": "https://betakit.com/toronto-founded-pagerduty-goes-public-sees-stocks-jump-in-market-debut/", "date": "Apr 12, 2019" } ], "launches": [] }, { "id": 560, "name": "StatusPage", "slug": "statuspage", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/c2e481e1e8141276646830bd6bd0c5bbf27ad892.png", "website": "http://statuspage.io", "all_locations": "San Francisco, CA, USA", "long_description": "Statuspage is the #1 status and incident communication tool, acquired by Atlassian in 2016. For more information, check out our new LinkedIn page: https://www.linkedin.com/showcase/18261232/", "one_liner": "The best way to communicate with users around downtime.", "team_size": null, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1367523759, "tags": [ "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2013", "status": "Acquired", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/statuspage", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2013/statuspage.json", "year_founded": 2012, "city": "San Francisco", "country": "US", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/statuspage", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/c2e481e1e8141276646830bd6bd0c5bbf27ad892.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 5525, "full_name": "Steve Klein", "title": "Founder", "founder_bio": "Previously co-founded Statuspage (S13) doing design, front-end dev, had hand in growth/support/success. Acquired by Atlassian in 2016 and transitioned to PM.\n\nAngel investing and working full-time w/ companies I invest in for a few months to help where I can.\n\nhttps://www.stevekle.in/ - stklein89@gmail.com", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/steveklein89/", "x_url": "https://twitter.com/stevenklein", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/8acaf9c0d90522bb93efd3a31710a25b8c307d78.jpg" }, { "user_id": 14879, "full_name": "Scott Klein", "title": "Founder", "founder_bio": null, "is_active": true, "linkedin_url": "https://www.linkedin.com/in/sbklein/", "x_url": null, "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/2de8da4ab3bd9e88d2fa0156e2f52f0cddbf44c6.jpg" }, { "user_id": 43661, "full_name": "Danny Olinsky", "title": "Founder", "founder_bio": null, "is_active": true, "linkedin_url": "https://www.linkedin.com/in/dannyolinsky", "x_url": null, "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/d7ef7d1c1e2b36955ef19d863c1ea961bec4b61f.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 690, "name": "Neptune.io", "slug": "neptune-io", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/282992bccaef3812d15f0bc6b45e2cce6670a8d0.png", "website": "http://neptune.io", "all_locations": "San Francisco, CA, USA", "long_description": "Neptune is an observability startup that offers one-stop incident enrichment and automation platform for IT Operations and DevOps teams.\r\n\r\nIt is a fact of life that servers and applications go down causing IT outages. Today most of the IT incident response is manual, tedious and lengthy resulting in more downtime and losses for businesses. At the same it's very painful for a DevOps engineer to wake up at midnight and on weekends to deal with server and application alerts. \r\n\r\nEnter Neptune, which lets a DevOps engineer automate his incident response steps --Troubleshooting + Resolution + Documentation, so that he/she can fix the incidents in minutes instead of hours. So it's a win-win deal : For businesses downtime is minimized, and for DevOps engineers they need not worry about a deluge of alerts and midnight wake up calls.\r\n\r\nBig companies like Amazon, Facebook etc built sophisticated incident response tools to achieve high availability and resiliency. We have architected and built one such tool for AWS. Now Neptune is making one such tool available for everyone. \r\n\r\nGive it a try at www.neptune.io", "one_liner": "A self-healing platform to enrich and fix server alerts automatically", "team_size": 11, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1398907889, "tags": [ "AIOps", "DevOps", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2014", "status": "Inactive", "industries": [ "B2B", "Infrastructure" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/neptune-io", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2014/neptune-io.json", "year_founded": 2013, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/neptune-io/", "x_url": "https://twitter.com/neptuneio", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/neptune-io", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/282992bccaef3812d15f0bc6b45e2cce6670a8d0.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 15542, "full_name": "Kiran Gollu", "title": "Founder/CEO", "founder_bio": "Co-founder/CEO at Oodle.ai\nPreviously, Head of Cloud Platform Engineering at Rubrik scaling from <5M to $600M ARR\nCofounder/CEO of Neptune.io, a YC backed observability startup\nFounding Engineer at AWS S3, DynamoDB\nAngel investor in enterprise startups tackling large markets", "is_active": false, "linkedin_url": "https://www.linkedin.com/in/kirangollu/", "x_url": "https://x.com/kkgollu", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/6f2ba555a7b4187e855886e48ea3d581bda608f0.jpg" }, { "user_id": 5519, "full_name": "Satish Talluri", "title": "Founder", "founder_bio": "Partner on Infra investing team at A16Z.\nProduct & Growth at AppDynamics\nCo-founder of Neptune.io (YC S14)\nManagement Consultant at BCG\nGPU design at Intel", "is_active": false, "linkedin_url": "https://www.linkedin.com/in/satishtalluri", "x_url": "https://x.com/satishtalluri", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/01cc980fe25849d9f6768df19659513b9531f236.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 909, "name": "Appcanary", "slug": "appcanary", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/812e3473cee76ddf3c226c05edc2883848c5fb14.png", "website": "http://appcanary.com", "all_locations": "Toronto, ON, Canada", "long_description": "Never run vulnerable software again.\r\nWe monitor your servers, and notify you when you need to update. \r\n\r\nTry us for free: https://appcanary.com", "one_liner": "Appcanary tracks the dependencies you use in your apps and servers…", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Security", "launched_at": 1430156121, "tags": [ "Security", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2015", "status": "Acquired", "industries": [ "B2B", "Security" ], "regions": [ "Canada", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/appcanary", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2015/appcanary.json", "year_founded": null, "city": "Toronto", "country": "CA", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/appcanary", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/812e3473cee76ddf3c226c05edc2883848c5fb14.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 73179, "full_name": "Filipa Mendonça-Vieira", "title": "Founder", "founder_bio": "perfect and upright, eschews evil. S15, acquired by GitHub.\n\npragmatic & empathetic engineering leader with a professional focus on security tooling and development at \"web-scale\".", "is_active": false, "linkedin_url": "https://ca.linkedin.com/in/phillmv", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/f583e7e461ab8fcde91bdd6b4487fb8df6671f0b.jpg" }, { "user_id": 68370, "full_name": "Max Veytsman", "title": "Founder", "founder_bio": "", "is_active": false, "linkedin_url": "https://www.linkedin.com/profile/view?id=116004429", "x_url": "https://twitter.com/mveytsman", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/c6819c0380fc97447df5f9c8d9f8a07925cdc8cf.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 22832, "name": "Parsagon", "slug": "parsagon", "former_names": [ "Parsagon", "Inc." ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/d7aefca6f08e98bcb987bd108cb1bbe99254dd91.png", "website": "https://parsagon.io", "all_locations": "San Francisco, CA, USA; Remote", "long_description": "Parsagon is an AI that can collect and analyze publications from any government body in any jurisdiction. It currently parses and analyzes legislation, press releases, speeches, meeting minutes, executive orders, and much more in 64 countries and counting.\r\n\r\nFortune/Global 500 companies, policy firms, and other organizations are using Parsagon to uncover policy insights in hundreds of jurisdictions that they would otherwise miss. Customers use us to:\r\n\r\n- Analyze changes in US state education policy\r\n- Generate reports on UK financial regulations\r\n- Monitor vaccine policies across the EU\r\n- And much more!", "one_liner": "AI to track and analyze public policy", "team_size": 2, "industry": "B2B", "subindustry": "B2B", "launched_at": 1614621550, "tags": [ "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2021", "status": "Active", "industries": [ "B2B" ], "regions": [ "United States of America", "America / Canada", "Remote", "Fully Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/parsagon", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2021/parsagon.json", "year_founded": 2020, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/parsagon/", "x_url": "https://twitter.com/Parsagon_io", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/d7aefca6f08e98bcb987bd108cb1bbe99254dd91.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 145104, "full_name": "Sandy Suh", "title": "Founder", "founder_bio": "Machine learning researcher specializing in program analysis and program synthesis. Using my experience to build an AI for public policy.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/sandy-suh-2ba05bb0/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/8afad87b2d165c28bfca4bb284282600e06150f2.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 98455, "title": "Parsagon: Universal API for Regulatory Data", "tagline": "Programmatic access to legislation, regulations, and more across 70+ countries", "body": "**TL;DR If you're scraping legislation, regulations, or other government publications, we provide a simpler option: a daily-updated dataset across 70+ countries, available via API. Check it out here: [https://parsagon.io/solutions/api](https://parsagon.io/solutions/api)**\n\nHi YC! We've built a global database of legislation, regulations, and other government publications. It updates daily and currently covers 70+ countries.\n\nWe originally built this for a handful of enterprise policy monitoring workflows, but we keep meeting founders who need reliable regulatory data for very different products/verticals (e.g., automating compliance workflows). Rather than trying to build every vertical ourselves, we're opening the dataset via API so you can build on top of it.\n\n**What It Looks Like**\n\nThe API is pretty simple: give a data category, and get a list of documents back:\n\n![uploaded image](/media/?type=post&id=98455&key=user_uploads/145104/f67f65dd-772a-4bfd-b1ee-a982b71afea4)\n\nYou can search for documents with date ranges and search queries as well:\n\n![uploaded image](/media/?type=post&id=98455&key=user_uploads/145104/c4c17ef5-5dfc-4d52-a468-33c3763099b3)\n\nTo see more examples, check out our docs: \n\n**What's Different About Parsagon**\n\nWe've seen a few solutions attempt to provide global regulatory/policy data, but they often are far from comprehensive. They often only include data for specific verticals/workflows, lack data for “smaller” jurisdictions, or just don't work as advertised (one of our customers said one provider only had half their data available via API - the other half was sent to them as excel spreadsheets of new regulations manually put together by humans each week).\n\nParsagon is built on AI research that generates and maintains data pipelines for government sources, which lets us scale coverage faster and keep it current.\n\nIf you need a source/jurisdiction we don’t have yet, just request it! We can usually add new coverage quickly. As one example, when a Global 500 tech company asked for \\~20 additional countries, we added them within two weeks.\n\n**Who Uses Our Data**\n\nParsagon helps Global 500 teams track new legislation, regulations, and policy developments. Examples:\n\n* A streaming platform tracks media/broadcasting rules in \\~60 countries\n* Multiple pharma companies track healthcare policy across 20+ countries\n* A major airline tracks policy developments across the EU\n\nIf you’re building anything that needs regulatory context (or you want to expand your existing product into more jurisdictions), we’d love to chat! You can email us at [**founders@parsagon.io**](mailto:founders@parsagon.io) or sign up at [**https://parsagon.io/solutions/api**](https://parsagon.io/solutions/api)", "url": "https://www.ycombinator.com/launches/Pbz-parsagon-universal-api-for-regulatory-data", "votes": 17, "created_at": "2026-03-04T18:02:22.640Z" }, { "id": 92447, "title": "Parsagon: AI to Track and Analyze Public Policy", "tagline": "Uncover policy insights you'd otherwise miss", "body": "Hi YC! I'm Sandy, and I'm trying to improve government transparency with Parsagon.\n\n**TL;DR**\n\nParsagon is an AI that can collect and analyze publications from any government body in any jurisdiction. It currently parses and analyzes legislation, press releases, speeches, meeting minutes, executive orders, and much more in 64 countries and counting.\n\n**Fortune/Global 500 companies, policy firms, non-profits, and other organizations are using Parsagon to uncover policy insights in hundreds of jurisdictions that they would otherwise miss.**\n\nAsk #1: If you want to try out Parsagon, email us at [founders@parsagon.io](mailto:founders@parsagon.io) or sign up for our beta at [parsagon.io](http://parsagon.io)\n\nAsk #2: We just launched on Product Hunt! We'd greatly appreciate if you could support [our launch](https://www.producthunt.com/products/parsagon)!\n\n## **😠The Problem:**\n\nGovernment data is much less accessible than it could be. Government affairs professionals spend countless hours sifting through legislation, press releases, speeches, etc. Tracking policy in one jurisdiction is hard enough, and any given organization may have to monitor dozens if not hundreds of different jurisdictions.\n\nCurrent platforms to track public policy are outdated. They track a limited set of sources, and they make you sift through those sources using clunky, keyword-based searches like the one below.\n\n![uploaded image](/media/?type=post&id=92447&key=user_uploads/145104/d2c61d43-51e6-465a-84df-124d0132a808)\n\n## **😃The Solution:**\n\nParsagon is an AI that can collect and analyze publications from any government body in any jurisdiction. You query Parsagon’s data in natural language, and Parsagon analyzes the policy developments relevant to you to create custom monitoring alerts, policy memos, and reports.\n\nhttps://www.youtube.com/watch?v=MnTf_Vz79AE\n\n# **🙏 Our Ask**\n\n* If you want to try out Parsagon, email us at [founders@parsagon.io](mailto:founders@parsagon.io) or sign up for our beta at [parsagon.io](http://parsagon.io)\n* We just launched on Product Hunt! We'd greatly appreciate if you could support [our launch](https://www.producthunt.com/products/parsagon)!", "url": "https://www.ycombinator.com/launches/O35-parsagon-ai-to-track-and-analyze-public-policy", "votes": 29, "created_at": "2025-07-25T15:08:03.272Z" }, { "id": 80386, "title": "Parsagon: Track business and policy developments", "tagline": "Use AI to monitor custom events", "body": "**Hi YC! We’re excited to announce that **[**Parsagon**](https://parsagon.io/)** can now monitor business and policy developments for you!**\n\n## **😠The Problem:**\n\nCurrent news monitoring tools are too noisy and too limited. They rely on simple keyword searches that fail to encompass complex topics, and they only monitor very standard news and social media sources. They lack the customization necessary to track many of the developments we care about in our industries and areas of expertise.\n\n## **😃The Solution:**\n\nParsagon’s AI monitors and analyzes events from sources across the web. You can specify custom sources to monitor, such as company websites, organizations’ meeting minutes, or government legislation. And you can filter for the events you care about using custom prompts.\n\n![uploaded image](/media/?type=post&id=80386&key=user_uploads/89898/1a881b41-a2c0-4c64-91b8-b673305ff667)\n\nFor example, you can use Parsagon to:\n\n* Collect data on mergers and acquisitions in the news\n* Get email alerts about announcements from state/local governments\n* Monitor bills related to a certain topic\n* And much more!\n\nJust tell Parsagon what types of events you want to monitor, and our AI will find what's relevant to you.\n\n## **How you can help:**\n\nIf you want to try out Parsagon, email us at [founders@parsagon.io](mailto:founders@parsagon.io) or sign up for our beta at [parsagon.io](https://parsagon.io/)", "url": "https://www.ycombinator.com/launches/KuY-parsagon-track-business-and-policy-developments", "votes": 36, "created_at": "2024-04-24T15:15:00.971Z" }, { "id": 79399, "title": "Parsagon - Real-time fundraising data for cheaper", "tagline": "A faster, cheaper way to get Crunchbase/Pitchbook data", "body": "**Hi YC! We’re excited to announce that **[**Parsagon**](https://parsagon.io/)** is offering real-time fundraising data similar to Crunchbase or Pitchbook, but for cheaper!**\n\n## **😠The Problem:**\n\nCollecting data on private companies is difficult and expensive, and as a result, if you want an API for fundraising data from Crunchbase or Pitchbook, you typically need to pay thousands of dollars per month.\n\n## **😃The Solution:**\n\nParsagon’s AI is now capable of auto-generating the scrapers and parsers needed to collect fundraising data, which allows us to offer the same data for much cheaper! To start, **we’re offering a real-time data feed starting at $50 per month!** You can use our API to monitor fundraising deals related to a given company, investor, industry, etc., and receive real-time data that looks something like this:\n\n```\n[\n {\n \"id\": 156,\n \"date\": \"2024-02-27\",\n \"round\": \"Series B\",\n \"amount_raised\": 43000000,\n \"valuation\": 500000000,\n \"valuation_type\": \"Post-money\",\n \"currency\": \"USD\",\n \"org\": {\n \"id\": 687,\n \"name\": \"Photoroom\",\n \"website\": \"https://www.photoroom.com\"\n },\n \"investors\": [\n {\n \"org\": {\n \"id\": 443,\n \"name\": \"Balderton\",\n \"website\": \"https://www.balderton.com\"\n },\n \"is_lead\": true\n },\n {\n \"org\": {\n \"id\": 688,\n \"name\": \"Aglae\",\n \"website\": \"https://aglaeventures.com\"\n },\n \"is_lead\": true\n },\n {\n \"org\": {\n \"id\": 689,\n \"name\": \"YCombinator\",\n \"website\": \"https://www.ycombinator.com\"\n },\n \"is_lead\": false\n }\n ],\n ...\n },\n ...\n]\n```\n\nWe’re also collecting historical fundraising data as well! Extensive coverage on that front will take some time, but **we’re currently collecting historical data on demand**. For example, one financial firm gave us a list of 1,000 companies on which they wanted historical data, and we were able to add that data to our API within a few days for a fraction of what it would cost from Pitchbook.\n\n## **How you can help:**\n\n* If you want to try out the API, email us at [founders@parsagon.io](mailto:founders@parsagon.io)\n* Do you know a VC or other financial firm that would be interested? Feel free to reach out to us at [founders@parsagon.io](mailto:founders@parsagon.io)\n* Not what the data you were trying to track? Reach out anyway! Our AI generated everything needed to get the data above in about two weeks, and it might be able to handle your use case just as easily!", "url": "https://www.ycombinator.com/launches/Ked-parsagon-real-time-fundraising-data-for-cheaper", "votes": 21, "created_at": "2024-03-14T20:13:13.386Z" }, { "id": 77241, "title": "Parsagon - Scrape sites with just their URLs", "tagline": "Give a list of URLs, get back a set of scrapers!", "body": "**Hi YC! We’re excited to announce that Parsagon now offers a new AI scraping service: you give a list of URLs, and you get back a set of scrapers!**\n\n**The Problem:**\n\nParsagon's AI allows you to generate web scrapers from natural language prompts. For example, you can get city population data from Wikipedia by telling our AI:\n\n> **_Go to _**[**_https://en.wikipedia.org/wiki/List_of_largest_cities_**](https://en.wikipedia.org/wiki/List_of_largest_cities)**_. Scrape data in the format \\[{\"city\": \"str\", \"link\": \"link\", \"population\": \"num\"}\\]_**\n\nThis works fine if you want to scrape a single site, but what if you want to scrape dozens, hundreds, or thousands of sites? What if each site works differently and needs a slightly different prompt? While it's faster than writing thousands of scrapers, writing thousands of prompts is still a lot of work.\n\n**The Solution:**\n\nNow you can give Parsagon a list of sites and the data you want from each site, and we'll pass that list to Upworkers trained to use Parsagon. The Upworkers write a prompt for each site, and you get scrapers for a fraction of the time/cost it would take to write code or hire developers! 🎉 For example, companies are using Parsagon to:\n\n* generate scrapers that monitor hundreds of state/local government websites for events and announcements related to economic policy\n* generate scrapers for thousands of company websites to get details on private companies’ financials and investors\n* scrape laws from government websites across the EU to create an AI legal assistant\n\n**The Ask:**\n\nIf you want to try out the service, email us at [**founders@parsagon.io**](mailto:founders@parsagon.io)!", "url": "https://www.ycombinator.com/launches/K5p-parsagon-scrape-sites-with-just-their-urls", "votes": 13, "created_at": "2024-01-08T21:15:26.083Z" }, { "id": 72447, "title": "Parsagon - GPT for browser automations ⚡️", "tagline": "Create browser automations with natural language", "body": "**Hi YC! We’re excited to announce that you can now use Parsagon to create browser automations with natural language!**\n\n**The Problem:**\n\nIf you want to do something like the following:\n\n* Go to a website, enter a query into a search bar, and get all the search results as JSON\n* Check a website for changes/updates every day\n* Test a form on your website with many different inputs, and run the tests on command\n\nYou usually have to write code. And then you might have to write code to monitor whether the website’s format has changed. And if the website changes, you need to write even more code. This is tedious and time-consuming for developers, and as an AI researcher, I wanted to use my research to build a better way to create automations…\n\n**The Solution:**\n\nWe’ve trained GPT to use Parsagon’s AI, which is built specifically to write Python code for web page interactions and web scraping. Now you can type something like:\n\n> _‘Go to _[_https://assessment.cot.tn.gov/TPAD/_](https://assessment.cot.tn.gov/TPAD/)_. Select \"Carter\" from the County dropdown. Type \"WEST DOE AVE\" into the Search Term box and hit enter. Scrape data in the format \\[{\"owner\": \"str\", \"address\": \"str\"}\\], clicking the Next button up to 3 times.’_\n\nand Parsagon will create a Python program to do what you described using Selenium. 🎉\n\n![uploaded image](/media/?type=post&id=72447&key=user_uploads/145104/52d57427-2220-4040-bc2e-5ac6c22a5a46)\n\nAs shown above, you can make complex browser automations in just 3 steps:\n\n* Type what you want the automation to do.\n* Parsagon will explore the website(s) you mentioned to create a Python program. Parsagon will show you an abridged version of what the program does, and if it looks correct, you can save the program to be run whenever you want. The program doesn’t use GPT, making it cheaper, faster, and more consistent.\n* You can use Parsagon to schedule your automations, alert you when a website’s format has changed and/or automatically try to adapt to format changes, and change the parameters of your automations (e.g., Parsagon will generate a program that allows you to change “Carter” and “WEST DOE AVE” to other values on each run in the example above).\n\n**Asks:**\n\n* If you want to try out the product, email us at [founders@parsagon.io](mailto:founders@parsagon.io) and we will move you up the waitlist to use the product.\n* If you know anyone who might be interested in using our tool, we’d appreciate an intro! Feel free to reach out to us at [founders@parsagon.io](mailto:founders@parsagon.io)", "url": "https://www.ycombinator.com/launches/IqV-parsagon-gpt-for-browser-automations", "votes": 19, "created_at": "2023-06-22T06:54:43.783Z" }, { "id": 70282, "title": "Parsagon: Universal API for product data", "tagline": "Collect product data from any website with one API call", "body": "### **✨ TL;DR**\n\nHi YC! We’re excited to announce that [Parsagon](https://parsagon.io/) is offering an API with which you can **collect product data from any e-commerce site with 1 API call**.\n\n### **🤯 The Problem**\n\nIf you’re trying to get data on products from e-commerce sites like Amazon, Etsy, etc., you essentially have 2 options:\n\n1. 1\\. Develop or purchase web scrapers for every site, which are time-consuming and expensive to build and maintain.\n2. 2\\. Use a product data API, where you provide a URL to a product page, and the API tries to extract product data from the page via ML models. This is cheaper, but these APIs often return inaccurate data for 20+% of product pages. We’ve seen these APIs say the price of a $350 TV was $16 because the warranty was $16, we’ve seen them say a price was $3400 when it was really $34.00, we’ve seen them say products were in stock when they were out of stock—the list goes on and on.\n\nAs developers, we at Parsagon have spent countless hours building and maintaining such web scrapers and wishing that we could use a product data API without sacrificing the accuracy and reliability of custom scrapers.\n\n### **💡 The Solution**\n\n**Parsagon provides a product data API with near 100% reliability.** We avoid the inaccuracies of previous product data APIs by using a novel AI to generate custom scrapers for every e-commerce website, ensuring that we consistently return accurate data for each site. Just give our API a URL for a product or list of products, and Parsagon will build a scraper for the site (if it hasn’t already) and return data for your product(s) with a standardized format.\n\nFor example, giving the API the URL for this page\n\n![uploaded image](https://bookface.ycombinator.com/media/?type=post&id=70282&key=user_uploads/145104/ac033de5-2e1e-4e30-b6bf-0b8dbb1faf28)\n\nyields:\n\n```\n{\n```\n\n```\n\"name\": \"Amazon eero 6 mesh Wi-Fi system | Supports speeds up to 500 mbps | Connect to Alexa | Coverage up to 4,500 sq. ft. | 3-pack, one router + two extenders, 2020 release\",\n\"price\": 179.99,\n\"regular_price\": 199.99,\n\"currency\": \"$\",\n\"sku\": \"B085WSCTS4\",\n\"out_of_stock\": false,\n. . .\n```\n\n```\n}\n```\n\nAnd giving the API the URL for this page\n\n![uploaded image](/media/?type=post&id=70282&key=user_uploads/145104/061755a7-0e01-4b75-8fa1-080259038282)\n\nyields\n\n> ```\n> [\n> \n> ```\n>\n> ```\n> {\n> \"name\": \"WICKER KURA BED- Pretty Stickers- Peel and Stick- Gifts- Adhesive Kura Bed Designs- Room Decor- Wicker Designs- Customizable Kura Beds\", \n> \"link\": \"https://www.etsy.com/listing/1141689620/wicker-kura-bed-pretty-stickers-peel-and?click_key=aa9bb9d8d3aaf77acefa1143c7d023c6631c1c59%3A1141689620&click_sum=5ea4b26e&ga_order=most_relevant&ga_search_type=all&ga_view_type=gallery&ga_search_query=beds&ref=search_grid-367299-1-1&pro=1&frs=1\", \n> \"price\": 58.99, \n> \"regular_price\": 78.66, \n> \"currency\": \"$\", \n> \"sku\": \"1141689620\", \n> \"out_of_stock\": false, \n> . . . \n> },\n> {\n> \"name\": \"Upholstered SOFT WALL PANELS for King or Queen size beds.\",\n> . . . \n> },\n> . . . \n> ]\n> \n> ```\n\n### **❓ Our Asks**\n\nIf you want an easier way to collect product data, email us at [founders@parsagon.io](mailto:founders@parsagon.io)", "url": "https://www.ycombinator.com/launches/IHa-parsagon-universal-api-for-product-data", "votes": 31, "created_at": "2023-03-14T19:55:07.287Z" }, { "id": 68413, "title": "Parsagon - Geographic data for half the price", "tagline": "Our AI collects land parcel data faster and cheaper than current solutions", "body": "**tl;dr: Parsagon is now offering a data service on all U.S. land parcels for 50% cheaper**\n\nParsagon is an AI that can auto-generate web scrapers and data pipelines. We had some customers ask for geographic data, so we decided to have Parsagon collect data on all land parcels in the U.S., as provided by municipal/county governments. We have collected data for several hundred counties within a few days, and we should have data covering the entire U.S. within a few weeks. We clean and standardize the data, which includes properties’ boundaries, coordinates, tax info, address info, data on structures, and [much more](https://parsagon.io/docs/parcels/parcel-data-fields).\n\nHow are we different from other data providers? Since we’re able to get the data much more easily, **we’ll charge you 50% less than what similar data providers are charging.** e.g., if competitors are charging $200/month for API access, we’ll charge $100.\n\n**Our ask:**\n\nIf you want access to the service, please contact us at [founders@parsagon.io](mailto:founders@parsagon.io)!", "url": "https://www.ycombinator.com/launches/HnR-parsagon-geographic-data-for-half-the-price", "votes": 13, "created_at": "2023-01-13T07:17:58.626Z" }, { "id": 66389, "title": "Parsagon: Real-time price indexes", "tagline": "Real-time price and inflation estimates via API", "body": "**tl;dr: Parsagon now offers real-time price indexes by scraping daily prices of retailers.** \n\nParsagon is an AI that can auto-generate web scrapers and data pipelines. With inflation affecting so many consumers and businesses, we had our AI generate dozens of scrapers to collect prices and track inflation for over 50,000 food products across 38 cities in North America. You can access daily inflation estimates per food category (e.g., bread, seafood) and per region (e.g., Northeast U.S., Western U.S.) via our API. For example, here’s the change in price for various foods in the U.S. over the past week:\n\n![uploaded image](/media/?type=post&id=66389&key=user_uploads/145104/6b268c1b-9eb6-4ca6-acd0-d7d19f356145)\n\nAnd earlier this month, when a jump in food inflation sent markets tumbling, our index reported that jump in inflation two weeks before the Bureau of Labor Statistics released the inflation numbers.\n\n**Our ask:**\n\nIf you want access to the API, please contact us at [founders@parsagon.io](mailto:founders@parsagon.io) for a free trial!\n\nWant prices for a different sector or industry? Let us know! We generated our scrapers for food in about a week, so we can likely get data for whatever industry you’re interested in. ", "url": "https://www.ycombinator.com/launches/HGn-parsagon-real-time-price-indexes", "votes": 17, "created_at": "2022-09-29T06:04:14.746Z" } ] }, { "id": 28074, "name": "Keep", "slug": "keep", "former_names": [ "Paladin" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/ca57c15ee63fa6b716d10d0f8f29a5e1732f44e4.png", "website": "https://www.keephq.dev", "all_locations": "Tel Aviv-Yafo, Tel Aviv District, Israel; Remote", "long_description": "Keep is an open-source AIOps platform. Keep acts as a single pane of glass to all your alerts from any monitoring tool and helps you turn 1000s into just 10s of meaningful alerts.\r\n\r\nKeep joined forces with Elastic on May 2025.", "one_liner": "The open-source AIOps platform", "team_size": 5, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1674164403, "tags": [ "SaaS", "B2B", "Open Source", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2023", "status": "Acquired", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "Israel", "Middle East and North Africa", "Remote", "Fully Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/keep", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2023/keep.json", "year_founded": 2023, "city": "Tel Aviv-Yafo", "country": "IL", "linkedin_url": "https://www.linkedin.com/company/keepalerting", "x_url": "https://twitter.com/keepalerting", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/ca57c15ee63fa6b716d10d0f8f29a5e1732f44e4.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Aaron Epstein", "url": "https://www.ycombinator.com/people/aaron-epstein" }, "founders": [ { "user_id": 1194329, "full_name": "Tal Borenstein", "title": "Founder", "founder_bio": "Formerly an engineering manager, now CEO. Tech-savvy with vast experience leading software teams of SaaS products.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/talbo", "x_url": "https://twitter.com/talboren", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/efce62171280879cdbed15a6c634d821fbab1bc9.jpg" }, { "user_id": 1196207, "full_name": "Shahar Glazner", "title": "Founder", "founder_bio": "Passionate about building stuff. Less is more. Long-distance runner.\nOccasionally (hopefully more) writes on shaharglazner.medium.com", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/shahar-glazner-29294654/", "x_url": null, "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/a59b17a4ebe14a0690f9119a56d1d26b7aaf9f17.jpg" } ], "jobs": [], "news": [ { "title": "\n\tElastic - Elastic Completes Acquisition of Keep\n", "url": "https://ir.elastic.co/news/news-details/2025/Elastic-Completes-Acquisition-of-Keep/default.aspx", "date": "May 20, 2025" }, { "title": "Elastic and Keep join forces to help users manage alerts and automate workflows | Elastic Blog", "url": "https://www.elastic.co/blog/elastic-and-keep-join-forces", "date": "May 06, 2025" }, { "title": "אחרי 8200 ו-Y Combinator: הסטארטאפ הישראלי Keep נחשף עם גיוס הון ראשון", "url": "https://www.geektime.co.il/keep-raises-seed-from-yc/", "date": "Oct 15, 2024" }, { "title": "Keep's AIOps platform helps ops teams reduce alert fatigue | TechCrunch", "url": "https://techcrunch.com/2024/10/15/keeps-aiops-platform-helps-ops-teams-reduce-alert-fatigue/", "date": "Oct 15, 2024" } ], "launches": [ { "id": 70094, "title": "Keep - Open source alerting CLI", "tagline": "Open source alerting CLI for all observability tools, with a simple and intuitive (GitHub actions-like) syntax.", "body": "Hey YC 👋, Shahar and Tal here 🚨\n\n![uploaded image](/media/?type=post&id=70094&key=user_uploads/1194329/3ba4c3ed-1965-4bf4-a261-985cecaff39e)\n\nWe are thrilled to share Keep with the YC community today!\n\n[Keep](https://github.com/keephq/keep) is an open-source alerting CLI alternative to Prometheus Alertmanager or Grafana Alerts, but for all observability tools, with a simple and intuitive (GitHub actions-like) syntax.\n\n**😫 Problem**\n\n* Engineering teams are frustrated with the explosion of tools that trigger alerts. On average, engineering teams have more than five tools that generate alerts.\n* Using proprietary vendor alerting/monitoring products locks you into the vendor, making it extremely challenging and time-consuming to migrate at a later stage.\n* Vendors today require you to send/integrate all your data, which is a big overhead when all you want to do is simple alerting.\n\n**🚨 Why Keep?**\n\n1. Simple and intuitive (GitHub actions-like) syntax.\n2. Declarative alerting that easily be managed and versioned in your source control and service repository.\n3. Alerts from multiple data sources for added context and insights.\n4. Freedom from vendor lock-in, making it easier to switch to a different observability tool if and when needed.\n\n💻 **Demo**\n\nPlay with Keep using our [Platform](https://platform.keephq.dev/)\\\n\\\n**🙏 Asks**\n\nTry out Keep and share your feedback on how it can be improved. Here’s our GitHub repo - \n\nHave any questions? Join our [Slack community](https://slack.keephq.dev/) and drop us a message", "url": "https://www.ycombinator.com/launches/IEY-keep-open-source-alerting-cli", "votes": 44, "created_at": "2023-03-09T17:20:58.252Z" } ] }, { "id": 28141, "name": "highlight.io", "slug": "highlight-io", "former_names": [ "Highlight" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/0d4301601e4f70936655aa2574a0bf0a798fe1bd.png", "website": "https://www.highlight.io/", "all_locations": "Seattle, WA, USA; Remote", "long_description": "highlight.io is the open source, full-stack monitoring platform. With just a few lines of code, Highlight makes it easy to understand what's slowing down your users and swiftly address errors and regressions on your web app. Imagine that you have a customer reporting an issue on your web app. With highlight.io, you can playback their exact interactions to reproduce the bug and get notified of any future errors that could happen.\r\n\r\nThat being said, we’re just getting started! We have plans to go way deeper into your observability stack to give you all of the visibility you need. Join us on this journey:\r\n- Star us on GitHub (https://github.com/highlight/highlight) to keep up with releases and news.\r\n- Join our Discord (https://highlight.io/community) to share product feedback and thoughts.\r\n", "one_liner": "The open source, fullstack monitoring platform: stop guessing why…", "team_size": 8, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1675367373, "tags": [ "Developer Tools", "SaaS", "B2B", "Open Source", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2023", "status": "Acquired", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada", "Remote", "Fully Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/highlight-io", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2023/highlight-io.json", "year_founded": 2023, "city": "Seattle", "country": "US", "linkedin_url": "https://www.linkedin.com/company/highlightrun", "x_url": "https://twitter.com/highlightio", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/0d4301601e4f70936655aa2574a0bf0a798fe1bd.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Jared Friedman", "url": "https://www.ycombinator.com/people/jared-friedman" }, "founders": [ { "user_id": 241768, "full_name": "Jay Khatri", "title": "Cofounder, CEO (acquired by LaunchDarkly)", "founder_bio": "I currently run the Observability business at LaunchDarkly. I was the CEO at Highlight, which was acquired by LaunchDarkly in March of 2025. ", "is_active": true, "linkedin_url": "https://linkedin.com/in/jay-khatri", "x_url": "https://twitter.com/theJayKhatri", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b9f1e7513e56139706e0a7fa29cc2cf7701527f0.jpg" }, { "user_id": 1250975, "full_name": "Vadim Korolik", "title": "Founder", "founder_bio": "Vadim Korolik is the founder and CTO of highlight.io, the open source session replay and error monitoring framework. A native of the Bay Area, Vadim previously worked at Pure Storage and drive.ai.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/vkorolik/", "x_url": "https://twitter.com/vkorolik", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/310d5870d701e4759b843740c85bf40bcf26f3e4.jpg" } ], "jobs": [], "news": [ { "title": "Highlight launches full-stack application monitoring platform, raises $8M | VentureBeat", "url": "https://venturebeat.com/programming-development/highlight-launches-full-stack-application-monitoring-platform-raises-8m/", "date": "Aug 15, 2023" } ], "launches": [ { "id": 68975, "title": "highlight.io - Open-source session replay & error monitoring", "tagline": "highlight.io is the open source, fullstack monitoring platform. We help engineers reproduce hard-to-crack issues, and understand how their product is used.", "body": "![uploaded image](/media/?type=post&id=68975&key=user_uploads/241768/1b79410b-8b41-4e1e-8869-c4ccfe7d2130)\n\n## **TL;DR:**\n\n[highlight.io](http://highlight.io) is an open-source **session replay and error monitoring** tool that helps teams reproduce hard-to-crack issues and understand product usage. We enable developers to understand WHY bugs are thrown on their web apps to eliminate the noisy alerts they get from more traditional monitoring tools.\n\n[highlight.io](http://highlight.io) is already being used in production by **2k+ developers and 80+ companies**. Get started for free at [app.highlight.io](http://app.highlight.io) and give us a ⭐️ at [github.com/highlight/highlight](http://github.com/highlight/highlight).\n\n### **The Problem**\n\nHow many times have you gotten an alert from Bugsnag or Sentry, and dismissed it because you weren’t sure about what caused it? And how many of you have an endless backlog of errors that your team _should_ triage, but haven’t gotten to yet?\n\nWe’ve found that when debugging an issue or customer request, **the toolset that you use is responsible for delivering one thing only: context**. And if there’s any gaps in this context, engineers are left in the dark.\n\nTo solve this, at our last companies, we found a lot of value in instrumenting a session replay-like tool, (similar to Logrocket or Fullstory), paired with an error monitoring tool (like Bugsnag). In fact, we’ve built these sorts of tools ourselves to hackily solve the problem. Enter [highlight.io](http://highlight.io).\n\n### **The Solution: Session Replay 🤝 Error Monitoring**\n\nWith a simple `yarn/npm` import in your frontend and an optional import in your backend, [highlight.io](http://highlight.io) gives you session replay and error monitoring out of the box. This means that you can:\n\n* Replay exactly what a user was seeing at the time of an issue/bug.\n* Reproduce the browser dev-tools for every user session, including the network requests, errors, and more.\n* Access all of the errors and exceptions (both frontend and backend) thrown in your app, paired with session replay.\n\nLastly, given the privacy-sensitive nature of the product, we have a very powerful privacy API that you can take a look at in our docs [here](https://www.highlight.io/docs/session-replay/privacy).\n\n## **💬 What’s Next?**\n\nWe’re just getting started! We have plans to go way deeper into your observability stack to give you ALL of the visibility you need; that means logs, traces and more! Join us on this journey:\n\n* ⭐️ us on [GitHub](https://github.com/highlight/highlight) to keep up with releases and news.\n* Join [our Discord](https://highlight.io/community)!", "url": "https://www.ycombinator.com/launches/HwV-highlight-io-open-source-session-replay-error-monitoring", "votes": 60, "created_at": "2023-02-02T19:04:50.853Z" } ] }, { "id": 28846, "name": "Parea", "slug": "parea", "former_names": [ "OptimusPrompt", "Parea AI", "Parea", "Parea AI" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/32972e7b6074f606373a477c496dcde4a5e81219.png", "website": "https://www.parea.ai", "all_locations": "New York City, NY, USA", "long_description": "Parea AI is the essential developer platform for debugging and monitoring every stage of LLM application development. We provide testing, evaluation, and monitoring in one unified platform. These capabilities help developers get visibility into LLM responses, quickly test and optimize prompts, and ensure customers get the best user experience.", "one_liner": "Test, evaluate & observe your LLM applications", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Operations", "launched_at": 1685551349, "tags": [ "Developer Tools", "Generative AI", "SaaS", "DevOps", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2023", "status": "Active", "industries": [ "B2B", "Operations" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/parea", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2023/parea.json", "year_founded": 2023, "city": "New York City", "country": "US", "linkedin_url": "https://www.linkedin.com/company/parea-ai", "x_url": "https://twitter.com/PareaAI", "fb_url": null, "cb_url": null, "github_url": "https://github.com/parea-ai", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/32972e7b6074f606373a477c496dcde4a5e81219.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 699950, "full_name": "Joel Alexander", "title": "Founder", "founder_bio": "Senior Software Engineer at Lyft.\nJ.P. Morgan TMT Investment Banking.\nBA @ Columbia University.\nMSc @ New York University", "is_active": true, "linkedin_url": "https://linkedin.com/in/ajoeljr", "x_url": "https://twitter.com/joel_a_wilde", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/7fdaebd022c79889e713b8b3d9e023ba6a0e1b65.jpg" }, { "user_id": 501039, "full_name": "Joschka Braun", "title": "Founder", "founder_bio": "Published my first paper in the highest-ranked number theory journal at age 17. Continued to pursue math until my M.Sc. from NYU as a Fulbright scholar. Noticed AI has more real world applications than math and started researching applications of deep learning to medical imaging at Covera Health. Built semantic search systems and automatic code generation software at Jina AI.\n\nhttps://joschkabraun.com", "is_active": true, "linkedin_url": "https://linkedin.com/in/joschkabraun", "x_url": "https://twitter.com/joschkabraun", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/9a835cbd9621cf7321af15246517e2417f3c0063.jpg" } ], "jobs": [], "news": [ { "title": "Y Combinator Demo Day 2023: Our favorites Day 2 | TechCrunch", "url": "https://techcrunch.com/2023/09/07/ycombinator-demo-day-two-summer-2023/", "date": "Sep 07, 2023" } ], "launches": [ { "id": 82145, "title": "Parea AI - Aligned & reliable LLM evaluations", "tagline": "Automatically create LLM-based evals aligned with human judgement", "body": "**tl;dr:** [Parea AI](https://www.parea.ai/) automates the creation of evals for your AI products. We achieve this by bootstrapping an evaluation function with human annotations. Allowing you to automagically turn “vibe checks” into scalable and reliable evaluations aligned with human judgment.\n\n# **😩 The Problem**\n\n![uploaded image](/media/?type=post&id=82145&key=user_uploads/501039/78024ed1-9aa2-4be8-a022-359d82b9ac59)\n\nEvaluating free-form text is often only possible by humans reviewing outputs or using LLMs to evaluate them. The former is laborious, slow, and expensive, while the latter often fails to evaluate the outputs correctly. For LLM evaluations to work properly, one needs to prompt engineer them; i.e., they require their own optimization process.\n\n# **🚀 The Solution**\n\n![uploaded image](/media/?type=post&id=82145&key=user_uploads/501039/b7c1f4c7-b654-40e6-9e1d-84ab93f4ad5f)\n\nThe best [LLM evals](https://joschkabraun.com) are adapted to your particular business use case & data. We've developed a method for uploading human annotations (via CSV or using our [Annotation Queue](https://docs.parea.ai/manual-review/queue)) and bootstrapping an evaluation to mimic the annotations. To create a human-aligned eval, you need as few as 20 sample annotations. Using your new LLM eval is as easy as copying the code into your codebase or using it directly via Parea's API. Check out [our docs](https://docs.parea.ai/manual-review/bootstrapped-eval) to see the complete workflow.\n\n# **🙏 Our Ask**\n\n* **Get started** on [**our free tier,**](https://app.parea.ai/sign-up) or [**book a chat**](https://calendly.com/parea-ai/chat) to discuss how we can help you evaluate your AI app!\n* **Support our** [**launch** **tweet**](https://x.com/JoschkaBraun/status/1812884216444510233) and follow us on [**LinkedIn**](https://www.linkedin.com/company/parea-ai/) and [**Twitter**](https://x.com/PareaAI)\n* **Share Parea** **AI** with anyone you know who is facing challenges [evaluating AI & LLM systems](https://joschkabraun.com).", "url": "https://www.ycombinator.com/launches/LMv-parea-ai-aligned-reliable-llm-evaluations", "votes": 8, "created_at": "2024-07-15T16:47:45.697Z" }, { "id": 73431, "title": "Parea AI – DataDog for LLM apps", "tagline": "Parea AI helps developers ideate, evaluate, and manage LLM-powered applications to consistently deliver quality user experiences.", "body": "**TL;DR:** [**Parea AI**](https://www.parea.ai/?utm_source=yc&utm_medium=post&utm_campaign=launch_yc&utm_content=tldr) is the essential developer platform for debugging and monitoring every stage of LLM application development. We provide debugging, testing, evaluation, and monitoring in one unified platform. These capabilities help developers get visibility into LLM responses, quickly test and optimize prompts, and ensure customers get the best user experience.\n\n![uploaded image](/media/?type=post&id=73431&key=user_uploads/699950/7cd2fa60-436e-46c0-b266-b88f3fd66b8c)\n\n### 🤔 **The Problem**\n\nLLMs hold infinite promise, but translating this potential into practical applications can be complex. Crafting precise prompts or building effective chains, tracking their performance, and optimizing them currently involves a disparate array of tools and a steep learning curve, leading to decreased productivity and increased time to market. This trial-and-error process can become taxing; in the words of one of our customers, **\"reading 1000s of words from ChatGPT outputs a day takes the fun out of building**.**\"**\n\n### 💡 **The Solution**\n\nParea was conceived to streamline this journey by simplifying the entire process of creating and managing LLM-powered systems. Our intuitive UI enables developers to **debug and test** their **prompts/chains**, offering deep visibility into the best-performing changes that dramatically improve LLM responses. The built-in Lab module lets you **compare various prompts, models, and parameter configurations side by side**, transforming rigorous testing into an insightful and fun task. Post-deployment, our robust **API ensures you have comprehensive analytics**, performance insights, and a complete tracking system for optimizing your application to excel in delivering user experiences.\n\n### 🚀 **Who is it for:**\n\nParea is for developers, AI enthusiasts, and organizations with a vision to leverage the power of LLMs. Whether you're a small team working on an LLM-driven product or a large company developing complex LLM-powered applications, **Parea simplifies the process and accelerates your time-to-productivity**.\n\n### 🎉 **Meet the Parea team**\n\nJoel and Joschka met in New York City and bonded over philosophy, learning, and of course, start-ups!\n\nJoel is a true Renaissance man. His quest has led him from a published microbiology researcher to International Politics at the Council of Foreign Relations, to learning Russian in Moscow, working on >$100B of M&A deals at J.P Morgan, and most recently honing his self-taught coding skills as a SWE at Lyft. With Parea AI, he's committed to enhancing the productivity of every AI builder.\n\n[Joschka](https://joschkabraun.com) is our Wunderkind. He published a paper in the highest-ranked Journal for Number Theory at age 17, graduated college at 19, took a gap year to do Kung Fu in China, then completed his Master's at NYU as a Fulbright scholar. Joschka spent the last 3 years on the AI/ML frontline doing computer vision at Covera Health, and semantic search & LLMs at Jina. Parea was born to tackle the issues he'd faced in debugging and building LLM applications like DevGPT.\n\n![uploaded image](/media/?type=post&id=73431&key=user_uploads/699950/4bdf0da2-6bfd-4139-8f0c-ea7e826fd062)\n\n### ❤️ **Our Ask**\n\nWe invite you to explore [Parea’s](https://www.parea.ai?utm_source=yc&utm_medium=post&utm_campaign=launch_yc2&utm_content=cta) capabilities for your LLM solution and join our welcoming community. Follow us on [Twitter](https://twitter.com/PareaAI) and on [LinkedIn](https://www.linkedin.com/company/parea-ai/).\n\nIf you have any questions or would like a personalized onboarding session for your team, write us at [founders@parea.ai](mailto:founders@parea.ai)", "url": "https://www.ycombinator.com/launches/J6N-parea-ai-datadog-for-llm-apps", "votes": 128, "created_at": "2023-07-26T03:34:53.972Z" } ] }, { "id": 29325, "name": "Respan", "slug": "respan", "former_names": [ "Keywords AI" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/511c79fc4e10857bff72dd38626234991f19c62c.png", "website": "https://respan.ai", "all_locations": "San Francisco, CA, USA", "long_description": "Respan gives teams a unified control plane to trace and evaluate agent behavior without guesswork, automatically surface issues, and fix what breaks faster. Respan connects production observability, automated and human-in-the-loop evaluations, and an adaptive AI gateway to close the loop between detection, decision, and action - so agents improve continuously in production.\r\n\r\nRespan is trusted by 100+ AI startups and enterprise teams. Today, the platform process 1B+ logs and 2T+ tokens every month, supporting more than 6.5M end users.", "one_liner": "Self-driving observability, evals, and gateway for AI agents", "team_size": 23, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1711136962, "tags": [ "AIOps", "Developer Tools", "SaaS", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2024", "status": "Active", "industries": [ "B2B", "Infrastructure" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": true, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/respan", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2024/respan.json", "year_founded": 2023, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/respan-ai", "x_url": "https://x.com/RespanAI", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/keywords-ai", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/511c79fc4e10857bff72dd38626234991f19c62c.png", "app_video_url": "https://bookface-images.s3.us-west-2.amazonaws.com/bf-public_profile/29382/app_video.mp4?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=ASIAQC4NIECAOTKCHR7A%2F20261020%2Fus-west-2%2Fs3%2Faws4_request&X-Amz-Date=20261020T022717Z&X-Amz-Expires=60&X-Amz-Security-Token=IQoJb3JpZ2luX2VjEHEaCXVzLXdlc3QtMiJHMEUCIQDeV7cQCYWa%2ByyakJaw7aH8e0HNWEyd2tO1Fn9KHjIpkgIgJPUveNomWrSf4Oq5DPuVDY7LKeaNJPvi5F2sbSx87Hoq5QMIOhAAGgwwMDYyMDE4MTEwNzIiDJ7%2FkqUU8p5mIeDC5CrCA7F4UggTYlVYiff4mYkoYiIpXa9r%2FKkUhcKwMNwtYpcuNTEBtH5xi9Hzkl3B%2FOYmUn%2FIacrdemsB9KrtOuRo%2FccaY%2Bs%2FSuIzime%2BcYuI7%2B%2Fg2nldjcm7YPCf32sIR8EflXc2916%2FDRMbztsRKDiQDL1ZKbK6YPHJppiQyVyRrrcCFD0X7tEf0D7QgK2DIpKnu8zGLzN4oAHn6xfHIgQnynrupcJbc9cga5AURtsRl7TfSM6XaGTtp1fNogyhUKX9f%2FYIL2pg03nlhLpRpDuqHpJ9qbbVsf%2FvboDHQ4IRRkiCaWH6GZwuTm%2FaXzTPIrAlVKH5%2F0RNaaExh%2Byw4VhHEEsSkHSs4mw%2BmNAKnW2FnWuldPRw5kn3j05dqhv5fkQxdp74JTYQe5wex1YN5PCok147bSKEuieYlIgH%2FiP9NY3Ym0ZRUWuo0zL40XWv9X3L1tLg%2BYWY0fCdhDKHGgTuesmbVajzEhu%2Bg9iWLptoK63cZ4UfO8oOd6sOc09XF6lBy%2BmD%2BSVmOfDWUOzz5ELD16slDTUwml0MZsttBoLDzldUc0olkl5nCAl23UqraiUXeaeW3yLOvJn%2BAD%2B%2FM6PvmZF0VzDn%2FdrWBjqlAVRsuTd0UsfPEqyKlBgBcSJTbWkHvSj10wzTCIqvOoZ3bz1TWza43iHLTCs7u%2FHmvoKE0qG5FTOsMYhLZb5XFdRretuo%2F1BjdoZt%2Fr0dHWGT4LLkmfYrkFclUq7u0pvMr6%2FR4I41ZYdCD3cWpGAcOrCQxKJukkkU%2Fx4O0IOnvZKvXzK4OISHe4XPgWQrnAJcSBOFX%2BXRwhpFbVXOOC%2BtEw%2BR3JTElw%3D%3D&X-Amz-SignedHeaders=host&X-Amz-Signature=28ca7223fbb3dd4e68c7e783631fa71ef9436c9af0c43239fdc74b5c7aa3fc85", "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/ece5e30ddc3a563c6c6cc623065ef6b3da426731.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/7cc1e99757189e637f88f49ba97e878857b4a0b7.png" ], "primary_partner": { "name": "Harj Taggar", "url": "https://www.ycombinator.com/people/harj-taggar" }, "founders": [ { "user_id": 927016, "full_name": "Andy Li", "title": "Founder/CEO", "founder_bio": "Founder @ Respan - AI gateway with built-in observability & automated evals.\n\nPreviously @ UIUC (dropped out) and Apple Product Design.", "is_active": true, "linkedin_url": "https://linkedin.com/in/hanheli", "x_url": "https://twitter.com/andydy42", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b58d188378e25a2e99ec8db052fe8f6f1d6df4fd.jpg" } ], "jobs": [ { "id": 92798, "title": "Founding Technical GTM", "url": "https://www.ycombinator.com/companies/respan/jobs/4ajQQf5-founding-technical-gtm", "location": "Alameda, CA, US / Remote (San Francisco, CA, US)", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$100K - $150K", "equity_range": "", "experience": "1+ years", "visa": "Will sponsor", "skills": [] }, { "id": 95364, "title": "Marketing", "url": "https://www.ycombinator.com/companies/respan/jobs/mpKAgcw-marketing", "location": "Alameda, CA, US / Remote (US)", "type": "Full-time", "role": "Marketing", "role_type": null, "salary_range": "$1 - $120K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 95461, "title": "Product Engineer", "url": "https://www.ycombinator.com/companies/respan/jobs/ATjyM7R-product-engineer", "location": "Alameda, CA, US / Remote (IN)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$1 - $200K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 95523, "title": "Senior Software Engineer, Frontend", "url": "https://www.ycombinator.com/companies/respan/jobs/ABWSs0E-senior-software-engineer-frontend", "location": "Alameda, CA, US / Remote (IN)", "type": "Full-time", "role": "Engineering", "role_type": "Frontend", "salary_range": "$1 - $200K", "equity_range": "", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 95524, "title": "Software Engineer, AI", "url": "https://www.ycombinator.com/companies/respan/jobs/9L4xvXt-software-engineer-ai", "location": "Alameda, CA, US / Remote (IN)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$1 - $200K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 95525, "title": "Software Engineer, Full-Stack", "url": "https://www.ycombinator.com/companies/respan/jobs/8s4G7k6-software-engineer-full-stack", "location": "Alameda, CA, US / Remote (IN)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$1 - $200K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 95526, "title": "Software Engineering Intern, Frontend (fall co-op)", "url": "https://www.ycombinator.com/companies/respan/jobs/apjciy1-software-engineering-intern-frontend-fall-co-op", "location": "Alameda, CA, US", "type": "Internship", "role": "Engineering", "role_type": "Frontend", "salary_range": "$3K - $6K / monthly", "equity_range": "", "experience": null, "visa": "US citizen/visa only", "skills": [] }, { "id": 95570, "title": "Chief of Staff", "url": "https://www.ycombinator.com/companies/respan/jobs/KaaccJS-chief-of-staff", "location": "Alameda, CA, US", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$1 - $250K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Respan raises $5M to bring proactive observability to AI agents - SiliconANGLE", "url": "https://siliconangle.com/2026/03/18/respan-raises-5m-bring-proactive-observability-ai-agents/", "date": "Mar 18, 2026" } ], "launches": [ { "id": 79874, "title": "Keywords AI - Unified DevOps platform to build AI applications", "tagline": "Dead simple way to deploy & monitor AI apps with 2 lines of code.", "body": "### **‼️ The Problem**\n\nLLM developers face many challenges that hinder the ability to quickly build and scale high-quality products that people love:\n\n* Time-consuming infrastructure setup and maintenance for deploying and scaling LLM applications\n* Difficulty collaborating on prompt engineering and evaluating models for specific use cases\n* Challenges monitoring and ensuring consistent performance in production for optimal user experience\n* Tedious process of creating datasets for fine-tuning and iteratively improving output quality\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/d44261ea-e426-4e84-9f5f-5e5c9e217417)\n\n\n### **🌟 Our Solution**\n\nKeywords AI is built for developers. It has every feature needed to build, deploy, and scale LLM applications:\n\n**Infrastructure and Deployment**\n\n* OpenAI-compatible integration with 2 lines of code\n* Supports 100+ LLMs with built-in load balancing, error handling, and dynamic routing\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/60b64a75-0987-4bbb-93e2-f946760abbee)\n\n\n**Model Playground**\n\n* Interactive playground for comparing prompts and A/B testing different models\n* Find the best LLM for your use case through experimentation and evaluation\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/bc21dd8f-740b-4600-8930-f887a86c81fa)\n\n\n**Prompt Management**\n\n* Intuitive prompt management system for versioning and deploying prompts\n* Collaborative workspace for teams to share and iterate on prompts\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/d943ff84-f908-4744-80b9-4b5dcafc4ac7)\n\n\n**Request Logging**\n\n* Request logs for tracing and debugging user-model interactions\n* Automated output evaluation for detecting hallucination and other quality issues\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/9761abd2-48c2-4dfd-92df-5f6b054f02ec)\n\n\n**Performance Evaluations**\n\n* Comprehensive evaluations to compare model performance with customized metrics\n* Monitor response quality over time to build product confidence\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/59f074ba-0d90-4eae-93f8-3fa2328d5750)\n\n\n**Usage Dashboard**\n\n* Real-time monitoring of model performance and user analytics\n* Comprehensive overview of product usage trends and user behavior\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/cc4903db-9d46-4c98-a44a-babe133fc848)\n\n\n**User Analytics**\n\n* Track growth, retention, and unit economics with detailed usage analytics\n* Sentiment analysis and topic/intent classification for user insights\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/eec04dbb-d665-4bee-8333-285de818a6bc)\n\n\n**Data Curation and Fine-Tuning**\n\n* ​Easily collect and curate production golden dataset for model optimization\n* One-click model fine-tuning for improved performance and cost savings​\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/d44261ea-e426-4e84-9f5f-5e5c9e217417)\n\n\n### **🤠 The Team**\n\n[Andy](https://linkedin.com/in/hanheli) and [Raymond](https://www.ycombinator.com/companies/www.pareon.com) both studied mechanical engineering at UIUC and have been best friends since college. \n\nAndy took a gap year to work as a product design engineer at Apple after freshman year, where he designed parts for the next-gen & next-next-next-gen AirPods. He dropped out after realizing his dream job wasn’t that fun.\n\nRaymond is so good at learning things. He graduated with the highest honor in just three years and self-mastered software engineering within a year. He dropped his robotics PhD offers from top schools.\n\nWe’ve worked on a bunch of very bad ideas together in the past year. Now we think we've stumbled upon a good one :)\n\n![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/b3ead780-3541-4b92-9e26-df04c3e891cc)\n\n![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/d44261ea-e426-4e84-9f5f-5e5c9e217417)\n\n### **🔥 Keywords AI is available today!**\n\n* [Get started free](https://keywordsai.co) with $15 in credits\n* [Chat with us](https://cal.com/keywordsai/demo) if you’re building with LLMs - we’d love to be helpful\n* Join our [Discord community](https://discord.com/invite/KEanfAafQQ) of AI builders\n* Follow us on [LinkedIn](https://www.linkedin.com/company/keywordsai/) and [X](https://twitter.com/keywordsai)\n\n ![uploaded image](/media/?type=post&id=79874&key=user_uploads/927016/d44261ea-e426-4e84-9f5f-5e5c9e217417)\n\n\n[team@keywordsai.co](mailto:team@keywordsai.co)", "url": "https://www.ycombinator.com/launches/KmI-keywords-ai-unified-devops-platform-to-build-ai-applications", "votes": 20, "created_at": "2024-04-02T14:24:37.708Z" } ] }, { "id": 29401, "name": "Raindrop", "slug": "raindrop", "former_names": [ "Sidekick [NOTE", "this is a duplicate app per request of Diana Hu. old id: zkoticha]", "Sidekick", "Invisible Tools", "Dawn", "dawn", "raindrop" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/53f3136371147d755b0a7f4891db765c11734499.png", "website": "https://www.raindrop.ai", "all_locations": "San Francisco, CA, USA", "long_description": "Monitor your AI agents the right way.\r\n\r\nAI engineers use Raindrop to get alerts about silent failures in their AI agents.\r\n\r\nRaindrop sends you alerts when your AI misbehaves and links straight to the events, so you can dig into the conversations or traces, understand the root cause, and fix it, fast.", "one_liner": "Sentry for AI Agents", "team_size": 9, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1710985617, "tags": [ "B2B", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2024", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Growth", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/raindrop", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2024/raindrop.json", "year_founded": 2023, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/raindrop-ai", "x_url": "https://x.com/raindrop_ai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/53f3136371147d755b0a7f4891db765c11734499.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Diana Hu", "url": "https://www.ycombinator.com/people/diana-hu" }, "founders": [ { "user_id": 1668573, "full_name": "Zubin Koticha", "title": "Founder/CEO", "founder_bio": "Building Raindrop – Sentry for AI products.\n\nPrev. cofounder & CEO of Opyn (acq. Coinbase), the first and largest DeFi options platform, with $15 + billion volume.\n\nUC Berkeley", "is_active": true, "linkedin_url": "https://linkedin.com/in/zkoticha", "x_url": "https://twitter.com/snarkyzk", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4ab33e3678516edb4a722ec477b49a48d86fe4c5.jpg" }, { "user_id": 309605, "full_name": "Alexis Gauba", "title": "Founder", "founder_bio": "Building Raindrop: Sentry for AI agents. \n\nPreviously Co-Founder at Opyn (acquired by Coinbase), the first and largest DeFi options platform ($15b+ volume), inventing a new financial asset class known as the power perpetual (an option that has no expiry). Dropped out of UC Berkeley EECS", "is_active": true, "linkedin_url": "https://linkedin.com/in/alexisgauba", "x_url": "https://twitter.com/AlexisGauba", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b11d9b354172a6ff1055340192988fdc431871f0.jpg" }, { "user_id": 1522257, "full_name": "Ben Hylak", "title": "Founder", "founder_bio": "building raindrop (sentry for ai products)\n\nI was previously on the Human Interface team at Apple for 4 years, building out visionOS. \n\nbefore that, dabbled with robotics + avionics", "is_active": true, "linkedin_url": "https://linkedin.com/in/benhylak", "x_url": "https://twitter.com/benhylak", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/04276ec80b6800d45fb6b2a91cbbe558d2d7894a.jpg" } ], "jobs": [ { "id": 65649, "title": "Product Engineer", "url": "https://www.ycombinator.com/companies/raindrop/jobs/3rxDvcR-product-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $250K", "equity_range": "0.10% - 1.50%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [ "Node.js", "Python", "TypeScript", "Torch/PyTorch", "Prompt Engineering" ] }, { "id": 86079, "title": "Backend Engineer", "url": "https://www.ycombinator.com/companies/raindrop/jobs/ogWQ7fP-backend-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Backend", "salary_range": "$150K - $250K", "equity_range": "0.10% - 1.50%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [ "Python", "TypeScript" ] }, { "id": 86080, "title": "ML Engineer", "url": "https://www.ycombinator.com/companies/raindrop/jobs/XJQYwdR-ml-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Machine learning", "salary_range": "$150K - $250K", "equity_range": "0.10% - 1.50%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [ "Python", "TypeScript", "Torch/PyTorch" ] }, { "id": 86172, "title": "Developer Experience Engineer", "url": "https://www.ycombinator.com/companies/raindrop/jobs/a4Vqw1I-developer-experience-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $250K", "equity_range": "0.10% - 1.50%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 92605, "title": "Founding Recruiter", "url": "https://www.ycombinator.com/companies/raindrop/jobs/EA7fuNw-founding-recruiter", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Recruiting & HR", "role_type": "Recruiter", "salary_range": "$150K - $200K", "equity_range": "0.10% - 0.50%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 97728, "title": "Forward Deployed Engineer (FDE)", "url": "https://www.ycombinator.com/companies/raindrop/jobs/jYbAqIG-forward-deployed-engineer-fde", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Devops", "salary_range": "$130K - $200K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 102883, "title": "Account Executive", "url": "https://www.ycombinator.com/companies/raindrop/jobs/bSz5bJR-account-executive", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$120K - $140K", "equity_range": "0.08% - 0.22%", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 103341, "title": "Sales Development Representative", "url": "https://www.ycombinator.com/companies/raindrop/jobs/7vLpHZX-sales-development-representative", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$70K - $110K", "equity_range": "0.05% - 0.15%", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Building the Safety Layer for AI Agents", "url": "https://www.youtube.com/watch?v=5XO7ZEOGpJc", "date": "Sep 17, 2026" }, { "title": "Raindrop raises $35M to monitor agents", "url": "https://www.axios.com/pro/enterprise-software-deals/2026/09/16/raindrop-crv-lightspeed-datadog", "date": "Sep 16, 2026" }, { "title": "Want to Know Why Your AI Agent Failed? There’s AI for That.", "url": "https://www.wsj.com/articles/want-to-know-why-your-ai-agent-failed-theres-ai-for-that-97a680be?gaa_at=eafs&gaa_n=AWEtsqdVmkMV2k7Sr_6KuR-BriI6nYI0WCuGV3kEnBFT-4QtLWIlapfZirpiHmgMeZ0%3D&gaa_ts=692e0171&gaa_sig=Xz1Zn2-He_brIjMQ9KK2JJUefbMrCjyW_AxqdIaMiylie5nJ5y12wKpTKjezNNlZrwKLcfsPZqocvU0KWcOtZw%3D%3D", "date": "Dec 01, 2025" } ], "launches": [ { "id": 91457, "title": "Raindrop Deep Search", "tagline": "Deep Research for your Production AI Data", "body": "https://www.youtube.com/watch?v=pN82WxN-_G0\n\nToday, we're excited to launch Raindrop Deep Search\n\nIt’s like Deep Research for your Production AI Data\n\nSearch for anything, and Raindrop automatically trains little models to accurately classify any topic or issue, across millions of events.\n\n𝗧𝗵𝗲 𝗣𝗿𝗼𝗯𝗹𝗲𝗺\n\nWe’ve heard from thousands of AI engineers and they’re struggling to track issues with their agents.\n\nImagine a user reports a problem: your agent is saying it can’t search the web for documentation. You need to know if this is a one-off problem or a much bigger issue… but how? Keyword search, or even semantic search, doesn’t tell the full story.\n\n𝗖𝗮𝗻’𝘁 𝘄𝗲 𝗷𝘂𝘀𝘁 𝘂𝘀𝗲 𝗧𝗿𝗮𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗘𝘃𝗮𝗹𝘀?\n\nOffline evals work well as unit tests. But since they’re running on preset data, you have no visibility into what’s actually happening in production.\n\nOnline evals just run these unit tests on a tiny sample of production data, leaving you blind to how widespread problems are.\n\n𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗶𝗻𝗴 𝗥𝗮𝗶𝗻𝗱𝗿𝗼𝗽 𝗗𝗲𝗲𝗽 𝗦𝗲𝗮𝗿𝗰𝗵\n\nThat’s why we built Deep Search. It’s like Deep Research for your production data.\n\nHow Deep Search works:\n\n1\\. Describe the issue (eg. agent failing to search the web)\n\n2\\. Deep Search finds examples out of millions of events\n\n3\\. Refine search with feedback\n\n4\\. Start tracking the issue\n\nDeep Search runs across all of your production data to give you an accurate metric of issue frequency.\n\nTraditional classification systems require humans to manually label thousands of data points. So to achieve this, Raindrop Deep Search introduces a new research breakthrough, bespoke few-shot classifiers, which only need a few examples.\n\nIt’s essentially bootstrapping weaker systems from stronger systems, ultimately training custom small models that analyze millions of events a day. You can think of it like creating materialized views for natural language.\n\nOnce you start tracking the issue you can use Raindrop to dive into traces and tool calls to find the root cause. And you can quickly confirm whether your fixes are effective by monitoring issue frequency and receiving real-time Slack alerts.\n\nYou can try out Deep Search at [raindrop.ai](http://raindrop.ai). \n\nWe’re excited to hear what you think!\n\n![uploaded image](/media/?type=post&id=91457&key=user_uploads/309605/b1eda0eb-8816-487f-a4f3-36bdaf02bb0b)\n\n", "url": "https://www.ycombinator.com/launches/Nn7-raindrop-deep-search", "votes": 10, "created_at": "2025-06-19T21:08:03.825Z" }, { "id": 89925, "title": "Raindrop - Sentry for AI Products", "tagline": " AI engineers use Raindrop to get alerts about hidden issues and wins in their AI products.", "body": "[https://youtu.be/IS7akU_3AnI](https://youtu.be/IS7akU_3AnI?feature=shared)\n\nAl products fail constantly—in ways both hilarious and terrifying.\n\nMeet **Raindrop**: the first Sentry-like monitoring platform for Al products.\n\n**The Problem**\n\nRegular software throws exceptions. But Al products fail silently.\n\nThese issues can be serious:\n\n* [Figma had to roll back their AI design product when it copied Apple’s designs, and they haven’t relaunched since](https://techcrunch.com/2024/07/02/figma-disables-its-ai-design-feature-that-appeared-to-be-ripping-off-apples-weather-app/?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAM5aRTBjXhPXfaoMBpoQxL9si-7R3qfaTR4r-n_mdzKu31uB6rUYGjE31z0H4aHVY1LlchyZ7r76Pufga45IMGKwFz0IxZ7eo-g2z1sYsta94AVsSj5CYjf6NKRd3m43VgSGLrG_IfdqMKtlFDuBvimRfLyF_aqb19vPgl-8B7eD)\n* [Air Canada got sued — and lost — when their chatbot incorrectly offered a customer a refund](https://www.forbes.com/sites/marisagarcia/2024/02/19/what-air-canada-lost-in-remarkable-lying-ai-chatbot-case/)\n* [Virgin Money came under fire when their chatbot scolded customers for using the company name, ‘Virgin’](https://www.latintimes.com/virgin-money-forced-apologize-after-chatbot-threatens-cut-off-bank-customers-who-use-word-573885#:\\~:text=Fintech%20commentator%20David%20Birch%20shared,reported%20by%20the%20Financial%20Times)\n\nDon’t make the same mistake.\n\nThe current status quo is sifting through millions of logs and trying debug flaky evals. Evals are straightforward - like unit tests, they confirm your model got specific test cases right. But in the real world AI chatbots and agents encounter millions of unpredictable actions. AI companies need powerful monitoring to discover production issues so they can create reliable AI products that don’t just pass tests, but genuinely resonate with users.\n\n**Solution**\n\nRaindrop sends you alerts when your AI misbehaves and links straight to the events, so you can dig into the conversations or traces, understand the root cause, and fix it, fast.\n\nDaily Alerts Include\n\n* **Issues**: detects issues like Assistant Forgetting, Laziness, Task Failure, User Frustration, and more depending on the type of app\n* **Wins**: surfaces what your product does well, so you can double down on those experiences and create evals\n\nThe Pro tier lets you go even further with:\n\n* **Custom Issues / Topics**: define and track any issue or topic\n* **Topic Clustering**: clusters data in real time to find your AI product’s most popular use cases, and what use cases have the most issues\n* **Signals**: finds patterns in explicit signals like thumbs up / thumbs down\n* **Deep Research:** deep research for your production data, letting you use natural language to search for any kind of behavior\n* **Traces**: track every step of your AI call\n* **Edge PII Redaction**: intelligently strip PII from any user and model messages\n* **Dataset Creation**: create custom datasets out of any set of events\n\nCompanies like [Clay.com](http://Clay.com), Tolans, Websim, and more have been using Raindrop to improve their AI products. They’ve been able to quickly iterate on fixes, see how issue incidence rates decrease in production, and confidently ship changes.\n\n“Raindrop has been invaluable as we've been growing quickly. It's critical for us to keep issue incidence below an acceptable threshold and become aware of any spikes. **It's like if we see an iOS crash report in Sentry, but for our AI capabilities.”** -Evan Goldschmidt, CTO Tolans\n\nLearn more about how Tolan uses Dawn [here](https://www.raindrop.ai/case-studies/tolans), and check out case studies with [New Computer](https://www.raindrop.ai/case-studies/new-computer) and [Unstuck AI](https://www.raindrop.ai/case-studies/unstuck).\n\nYou can get started [here](https://www.raindrop.ai), excited to hear what you think :)\n\n![uploaded image](/media/?type=post&id=89925&key=user_uploads/309605/d77c5ea9-2b73-4455-96a6-db02c319651c)\n\n", "url": "https://www.ycombinator.com/launches/NOP-raindrop-sentry-for-ai-products", "votes": 11, "created_at": "2025-04-28T20:25:57.382Z" }, { "id": 79667, "title": "Dawn - Analytics for AI products", "tagline": "We transform user requests and model outputs into metrics you actually care about.", "body": "Hey everyone!\\\n\\\nI’m [Zubin](https://twitter.com/snarkyzk), from dawn. Excited to share what we’ve (me, [Ben](https://twitter.com/benhylak), & [Alexis](https://twitter.com/alexisgauba)) been building.\n\n[Dawn](https://www.dawnai.com/) is analytics for AI products. We transform user inputs and model outputs into metrics you care about, letting you make meaning out of billions of tokens.\n\n**AI Companies love us for**\n\n* User Segmentation: automatically group users by behavior\n* Understanding Churn: see how requests (or responses) cause users to churn\n* Categorizing Search Requests: tag user requests + use categories in-app\n* Content Moderation: identify + remove harmful content\n* Flagging Bad Responses: track user requests that are vague, don’t match the source language, or anything else\n* Track any use case: discover categories from unstructured data\n\n_(It's not just for Al companies. Dawn can categorize just about anything! — _[_contact us_](https://wv6y9hc66q4.typeform.com/to/qHmrVeA3)_ to chat)_\n\n**How it Works**\n\n1. **3 lines to integrate:** with one line of code, you can start sending user data from your app to dawn (and even get back categorizations!)\n2. **Define categories:** Dawn automatically discovers categories from your raw data. You can also define your own. You can delete, edit, reorganize, and create new categories at any time.\n3. **Self-improving, on the fly:** Dawn constantly gets better at understanding your data. 
A simple feedback interface gives your team visibility and control over categorization.\n4. **Connects to your existing analytics:** we have plugins for Snowflake, Mixpanel, Slack, Amplitude, and more -- so you can track whatever you want, wherever you want.\n\n\\*\\*What early users say about us:\\*\\*\\\n“I’m feeling a total a-ha moment with this. Truly cool to see.” - Gabriel Birnbaum, Can of Soup\\\n\\\n**Let’s work together**\\\nReach out at [zk@dawnai.com](mailto:zk@dawnai.com), or sign up for a demo [here](https://wv6y9hc66q4.typeform.com/to/qHmrVeA3). We’d love to hear from you!\n\n![uploaded image](/media/?type=post&id=79667&key=user_uploads/1668573/86bc4a34-fc8f-469a-9d07-de695eb11bc3)\n\n", "url": "https://www.ycombinator.com/launches/Kix-dawn-analytics-for-ai-products", "votes": 21, "created_at": "2024-03-25T15:28:55.097Z" } ] }, { "id": 29524, "name": "Coval", "slug": "coval", "former_names": [ "Datawave" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/adc31ffe74137f2876c08ce7153b3a5cfd5001ca.png", "website": "https://coval.dev", "all_locations": "San Francisco, CA, USA", "long_description": "Coval is a simulation & evaluation platform for autonomous AI agents, helping engineers launch dependable assistants across chat, voice, and other modalities. \r\nWe simulate thousands of scenarios engineers don't have to manually test their agents. Our CI/CD evaluations automatically simulate and detect regressions.", "one_liner": "Simulation & Evaluation that scales voice and chat AI agents", "team_size": 10, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1725341816, "tags": [ "Developer Tools", "SaaS", "B2B", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Summer 2024", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/coval", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2024/coval.json", "year_founded": 2024, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/covaldev", "x_url": "https://x.com/covaldev", "fb_url": null, "cb_url": null, "github_url": "https://github.com/coval-ai", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/adc31ffe74137f2876c08ce7153b3a5cfd5001ca.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/bea8c9d71c2d41cbf4ea90c5192dbfc49f44438d.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/e1dc8955899eae685e3429228f0a731ed7402b34.png" ], "primary_partner": { "name": "Harj Taggar", "url": "https://www.ycombinator.com/people/harj-taggar" }, "founders": [ { "user_id": 1019731, "full_name": "Brooke Hopkins", "title": "Founder", "founder_bio": "Building evaluation for AI agents, previously led eval infra at Waymo 🚙", "is_active": true, "linkedin_url": "https://linkedin.com/in/bnhop", "x_url": "https://twitter.com/bnicholehopkins", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/0577036f83dc9c8cd80e528eaf8740bdbf2cd06f.jpg" } ], "jobs": [ { "id": 92828, "title": "Account Executive", "url": "https://www.ycombinator.com/companies/coval/jobs/sFLtjLc-account-executive", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$110K - $350K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 92854, "title": "Founding Product Engineer", "url": "https://www.ycombinator.com/companies/coval/jobs/ILkSK7r-founding-product-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$100K - $300K", "equity_range": "0.01% - 1.00%", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [ "Amazon Web Services (AWS)", "Python", "React", "Distributed Systems", "Data Modeling" ] }, { "id": 92860, "title": "Forward Deployed Engineer", "url": "https://www.ycombinator.com/companies/coval/jobs/ITTPluu-forward-deployed-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$100K - $250K", "equity_range": "0.01% - 1.00%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [ "Amazon Web Services (AWS)", "Python", "React", "TypeScript", "Next.js" ] }, { "id": 110904, "title": "Deployment Strategist", "url": "https://www.ycombinator.com/companies/coval/jobs/OcJmxbg-deployment-strategist", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Product", "role_type": null, "salary_range": "$150K - $250K", "equity_range": "", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 115250, "title": "Business Development Representative ", "url": "https://www.ycombinator.com/companies/coval/jobs/KAYIxhg-business-development-representative", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$100K - $140K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 116078, "title": "Solutions Engineer", "url": "https://www.ycombinator.com/companies/coval/jobs/tZSJGPI-solutions-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$180K - $250K", "equity_range": "", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "What It Actually Takes to Deploy a Voice Agent to a Fortune 500", "url": "https://www.youtube.com/watch?v=eSm_9tb5ZbY", "date": "Jun 24, 2026" }, { "title": "Coval raises $28M to make better voice agents", "url": "https://www.axios.com/pro/enterprise-software-deals/2026/06/24/coval-ai-voice-agents-enterprise", "date": "Jun 24, 2026" }, { "title": "Coval evaluates AI voice and chat agents like self-driving cars | TechCrunch", "url": "https://techcrunch.com/2025/01/23/coval-evaluates-ai-voice-and-chat-agents-like-self-driving-cars/", "date": "Jan 23, 2025" } ], "launches": [ { "id": 83700, "title": "Coval - Simulation & Evaluation for AI Agents", "tagline": "Ship AI agents faster with a simulation and evaluation platform for chat and voice assistants.", "body": "Teams are racing to market with AI agents, but slow manual testing processes are holding them back. Engineers currently spend hours manually evaluating and playing whack-a-mole just to discover that fixing one issue introduces another.\n\nAt Coval, we build automated simulation and evaluation for AI agents inspired by the autonomous vehicle industry to boost test coverage, speed up development, and validate consistent performance.\n\nWe have a waitlist, but YC companies go first! Grab some time here: \n\n![uploaded image](/media/?type=post&id=83700&key=user_uploads/1019731/e39b0b4e-22ea-4e62-85a4-d5421cafb7d5)\n\n# **Our Story**\n\nHey! I’m [Brooke](https://www.linkedin.com/in/bnhop/), the founder of [Coval](https://www.coval.dev/). 👋\n\nBefore starting Coval, I led the evaluation job infrastructure team at [Waymo](https://waymo.com/). I coded the first versions of our dataset storage and other foundational simulation systems, and my team built all of the dev tools for launching and running evals.\n\nThrough my conversations with hundreds of engineering teams at startups and enterprises, I've seen that AI agents—models that operate independently and handle complex tasks—are facing similar challenges to those in self-driving. \n\nIn the early days, autonomous vehicle companies relied heavily on manual evaluation, testing the self-driving cars on racetracks and city streets (remember when autonomous cars still had safety drivers?). However, as startups scaled, a significant shift happened: we moved towards simulating every code change in a “virtual” environment, using the vast amounts of data we collected. The new approach dramatically improved vehicle behavior, leading to hundreds of autonomous cars zipping around the San Francisco streets today!\n\nThis story mirrors what's happening today with AI agents across various industries. Teams are coming up with promising prototypes but often hit a wall when it comes to their reliability. \n\nAs we build for the future, where AI agents execute much of our work, ranging from sending emails to prescribing medication, the risks posed by untested systems could severely throttle the progress.\n\nAt Waymo, I developed tools that tested each code modification made by engineers, ensuring that every change improved the Waymo Driver's performance. I believe this methodical approach was key in helping our team address edge cases and maintain peak performance, and it ultimately cemented Waymo's status as a leader in the autonomous vehicle space.\n\nNow, at Coval, we’re taking these proven strategies and adapting them in a completely new way to speed up the development of AI agents. Our goal is to help engineers build agent experiences that genuinely work for users in the real world.\n\nAutomated simulation and evaluation are critical to trusting agents with impactful tasks across industries.\n\n# **Working With Us**\n\nBuilding agents or know someone? Let’s talk! [Grab time with me here](https://bit.ly/coval-demo) for a quick intro, or message me at [brooke@coval.dev](mailto:brooke@coval.dev).", "url": "https://www.ycombinator.com/launches/Lm0-coval-simulation-evaluation-for-ai-agents", "votes": 81, "created_at": "2024-09-03T12:43:30.586Z" } ] }, { "id": 30403, "name": "Aravolta", "slug": "aravolta", "former_names": [ "CentralAxis" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/fef90b2fedf3b6bcd07d770648c6074b9717d2a0.png", "website": "https://www.aravolta.com/", "all_locations": "San Francisco, CA, USA", "long_description": "Aravolta provides a single pane of glass software for data centers to increase uptime, reduce costs, and improve performance.\r\n\r\nA single utility node onboards every asset within an hour and detects hardware failures, hotspots, and other issues.\r\n\r\nIt also generates reports on important metrics like Power Usage Effectiveness (PUE), energy metering, resource utilization, and much more. ", "one_liner": "Software to monitor, control, and optimize data centers.", "team_size": 4, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1746569409, "tags": [ "Enterprise Software", "Cloud Computing", "Monitoring" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Spring 2025", "status": "Active", "industries": [ "B2B", "Infrastructure" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/aravolta", "api": "https://devasheeshg.github.io/yc-api/batches/spring-2025/aravolta.json", "year_founded": 2024, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/centralaxis", "x_url": "https://x.com/CentralAxis_X25", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/fef90b2fedf3b6bcd07d770648c6074b9717d2a0.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Nicolas Dessaigne", "url": "https://www.ycombinator.com/people/nicolas-dessaigne" }, "founders": [ { "user_id": 1833246, "full_name": "Margarita Groisman", "title": "Founder/CEO", "founder_bio": "Worked on data center deployments at Microsoft, now my company Aravolta builds software for data centers. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/margarita-groisman/", "x_url": "https://x.com/Marg_Groisman", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/5150584412ea3eb0182f8e63c266e0f281c95f7c.jpg" }, { "user_id": 2213346, "full_name": "Jack Sutton", "title": "Founder/CTO", "founder_bio": "Lifelong software engineer", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/sutton-jack", "x_url": "https://x.com/JackSutton_SF", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/8af4649e155b78aa37df060b30c652788ea65252.jpg" } ], "jobs": [ { "id": 76973, "title": "Senior Software Engineer", "url": "https://www.ycombinator.com/companies/aravolta/jobs/5mXBTwN-senior-software-engineer", "location": "San Francisco, CA, US / Remote (US)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$140K - $200K", "equity_range": "0.40%", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [ "Bash/Shell", "Node.js", "PostgreSQL", "TypeScript", "Next.js", "Tailwind" ] }, { "id": 80505, "title": "Intern", "url": "https://www.ycombinator.com/companies/aravolta/jobs/xNqAuCd-intern", "location": "San Francisco, CA, US / Remote (US)", "type": "Internship", "role": "Engineering", "role_type": "Full stack", "salary_range": "$4K - $15K / monthly", "equity_range": "", "experience": null, "visa": "US citizen/visa only", "skills": [ "Amazon Web Services (AWS)", "Node.js", "React", "TypeScript" ] }, { "id": 92973, "title": "Forward Deployed Engineer", "url": "https://www.ycombinator.com/companies/aravolta/jobs/F5owCzV-forward-deployed-engineer", "location": "San Francisco, CA, US / Remote (US)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $240K", "equity_range": "0.25% - 0.50%", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 111389, "title": "Sales Development Representative", "url": "https://www.ycombinator.com/companies/aravolta/jobs/ySYYxBA-sales-development-representative", "location": "San Francisco, CA, US / Remote (US)", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$80K - $200K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] } ], "news": [], "launches": [] }, { "id": 30863, "name": "The Context Company", "slug": "the-context-company", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/897cd2d31bc7dc869e170422ebc8e24a66f3e2bc.png", "website": "https://www.thecontextcompany.com", "all_locations": "San Francisco, CA, USA", "long_description": "We analyze every conversation between your customers and your AI agent to find what's losing you customers and money. Then we help ship the fix and prove it worked.\r\n\r\nNobody on the team can read every conversation once an agent is in production. TCC does, and turns them into patterns tied to the accounts they affect: where customers get stuck, what they wish the agent could do, and where it's spending more than it needs to.\r\n\r\nEngineers ask about a pattern from their IDE through our MCP, get the conversations and traces behind it down to the tool call or prompt that went wrong, ship the fix, and benchmark cost and quality before and after. Product sees what customers are actually trying to do with the agent and where it falls short, broken out by topic and account, which becomes the backlog for what to build next. Customer success sees which accounts the agent is struggling with before the customer says anything, and steps in with the full conversation on hand.\r\n\r\nAny agent framework, a few lines of code. If you're running an agent in production or about to launch, ping us.", "one_liner": "Find what's losing you customers and money in your AI agent and fix it", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1759953898, "tags": [ "Developer Tools", "B2B", "Analytics", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Fall 2025", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/the-context-company", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2025/the-context-company.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/the-context-company/", "x_url": "https://x.com/thecontextco", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/897cd2d31bc7dc869e170422ebc8e24a66f3e2bc.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/f79a5b20add310d8c52a8e37b33c10a81faf8f5e.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/af950c1fa4e85da46b706668594f2e3898e07954.png" ], "primary_partner": { "name": "Nicolas Dessaigne", "url": "https://www.ycombinator.com/people/nicolas-dessaigne" }, "founders": [ { "user_id": 2357695, "full_name": "Arman Kumaraswamy", "title": "Founder", "founder_bio": "Cofounder at The Context Company (Monitoring for AI agents)\n\nI've been in love with startups since I was in high school :)\n\nPrev studied CS+Math at UF, and engineered at Mintlify, Apten, RevisionDojo, and Solace Health", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/armankumaraswamy/", "x_url": "https://x.com/ksw_arman", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/8401e71daab0e1115cff57cc14f4453c2bce6052.jpg" }, { "user_id": 2579624, "full_name": "Rohil Agarwal", "title": "Founder", "founder_bio": "What is up! Previously studied EECS + Business at UC Berkeley M.E.T. and worked on Gmail Intelligence at Google, where I built models powering intent detection for 26B+ daily emails.\n\nNow building The Context Company, where we are powering developers to ship AI agents they can trust. I'm excited about solving context engineering 🙃\n\nOh, and I take one selfie every day!", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/rohil-ag/", "x_url": "https://x.com/rohil_ag", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4d30f11084359006ce25e5cf6c97ee26d09cb543.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 94947, "title": "The Context Company - Catch Silent AI Failures Before They Break Production", "tagline": "Full observability for AI agents, from bad tool calls, loops, and hallucinations to latency and cost, in 10 lines of code.", "body": "What is up! We’re [Arman](https://www.linkedin.com/in/armankumaraswamy/) and [Rohil](https://www.linkedin.com/in/rohil-ag/) - cofounders of [The Context Company](https://thecontextcompany.com)\n\n**TL;DR**\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/129e5656-94c6-42e6-a2ba-6ac545b0ef94)\n\nSetup takes under 10 lines of code, and we now support Vercel AI SDK, LangChain, and LangGraph, with more frameworks coming fast.\n\n**Our Ask**\n\n* If you’re building with Vercel AI SDK, LangChain, or LangGraph, use The Context Company for observability.\n* Setup takes just a few lines - DM us on Bookface or Slack, or email [founders@thecontextcompany.com](https://mailto:founders@thecontextcompany.com), and we’ll personally onboard you in 15 minutes.\n* Not using those frameworks yet? Tell us what you’re building with, and we’ll add support fast.\n\n\n\n**The Problem**\n\nAI agents often fail quietly. Traditional observability tools only show latency or stack traces. They miss the silent layer of failures that happen inside reasoning steps or tool calls.\n\nMost teams end up waiting until users report issues, and by then, it’s already too late. As a developer, I know you’ve been there scrolling through endless logs, trying to piece together what an agent actually did and why it failed (because trust us, we have too 🙏🏽).\n\n**We’re rethinking observability from the ground up and making silent failures a first-class primitive in agent monitoring.**\n\n\\*\\*Our Solution:\\*\\*\\\nInstead of just visualizing traces, we analyze them and detect reasoning breakdowns across runs. In addition to traditional metrics like latency, cost, and token usage, The Context Company surfaces failures in a clean, developer-friendly dashboard.\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/861e36d0-5a8b-4555-a738-e2489729f99f)\n\nAnd we’ve gone beyond the cloud:\n\nYou can now run our developer package locally to debug traces on your machine, even when you’re offline or iterating fast.\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/b7722f3f-a018-4c46-806c-7ef4be35bd44)\n\n**Our Progress**\n\n* Framework support: Vercel AI SDK, LangChain, and LangGraph.\n* Instant setup: <10 lines of code - no refactoring or modifying your agent needed. This is what it looks like:\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/560baa96-5818-4d20-bf82-43f2d3775ca3)\n\n* Local debugging: Run observability locally on your laptop for real-time trace analysis.\n* Dashboards: Full visibility into agent runs, tool calls, latency, costs, and semantic failures.\n\n**The Team**\\\nArman and Rohil have been best friends since 6th grade (8+ years now!). \n\nArman optimized the public-facing AI agent at Mintlify, where he cut failure rates by 82%, and worked with enterprise design partners to build custom agents. Rohil worked on Gmail Intelligence at Google, reducing hallucinations in intent detection by 27% across 26B+ daily emails. \n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/70e7935c-7606-4e8f-a7af-4ebf44cf5514)\n\nhttps://www.thecontextcompany.com\n\nP.S. Even if you’re not using AI SDK, LangChain, or LangGraph, we’re adding support for other frameworks soon. Tell us what you need 😎\\\n\\\n\\\nDashboard:\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/e9425baf-3e5f-4098-b64f-497a74232f5a)\n\n![uploaded image](/media/?type=post&id=94947&key=user_uploads/2579624/ece01a4a-5089-46dd-a12a-eeef441f5cd3)\n\n", "url": "https://www.ycombinator.com/launches/OhP-the-context-company-catch-silent-ai-failures-before-they-break-production", "votes": 34, "created_at": "2025-10-27T21:36:21.070Z" } ] }, { "id": 30969, "name": "Lemma", "slug": "uselemma", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/335dca5772b36d447d2e1c9b0f730539813d1846.png", "website": "https://www.uselemma.ai/", "all_locations": "San Francisco, CA, USA", "long_description": "Lemma is production monitoring for AI agents. We catch the silent failures your observability tools and evals miss (think bad tool calls, lost context, and infinite loops) before your users find them.", "one_liner": "Production Monitoring for AI agents", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1761703288, "tags": [ "Developer Tools", "Analytics", "Infrastructure", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Fall 2025", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/uselemma", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2025/uselemma.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/uselemma", "x_url": "https://x.com/uselemma_ai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/335dca5772b36d447d2e1c9b0f730539813d1846.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Nicolas Dessaigne", "url": "https://www.ycombinator.com/people/nicolas-dessaigne" }, "founders": [ { "user_id": 1798805, "full_name": "Jerry Zhang", "title": "Founder/CEO", "founder_bio": "Co-founder @ Lemma (YC F25) | Reliability platform for AI agents", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/jerry-n-zhang/", "x_url": "https://x.com/zjearbear", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/5ea73d82ad13ecb85985095e699c5a6b06cff636.jpg" }, { "user_id": 1374781, "full_name": "Cole Gawin", "title": "Founder/CTO", "founder_bio": "developing the next generation of intelligent systems @ lemma (f25) | ex-USC", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/colegawin", "x_url": "https://x.com/colegawin_", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/deca21906919edc6e42060d0665ec4c3ae3d597a.jpg" } ], "jobs": [ { "id": 104341, "title": "Product Engineer", "url": "https://www.ycombinator.com/companies/uselemma/jobs/wrFU6bU-product-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$120K - $200K", "equity_range": "0.50% - 2.00%", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 113900, "title": "AI/ML Engineer", "url": "https://www.ycombinator.com/companies/uselemma/jobs/vyHNWKs-ai-ml-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "AI engineer", "salary_range": "$120K - $200K", "equity_range": "0.50% - 2.00%", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] } ], "news": [], "launches": [ { "id": 95219, "title": "Lemma: Continuous Learning for AI Agents", "tagline": "We enable AI agents to continuously improve by turning real user feedback into automated prompt optimizations. ", "body": "**TL;DR:**\n\nLemma is the first evaluation + observability platform built not just to measure performance, but to improve it automatically. We help AI agents learn from real user feedback and production data, closing the loop so your prompts and agents continuously optimize themselves over time.\n\n**Launch Video:** https://www.youtube.com/watch?v=E4_v-pY_4fs\n\nHey everyone! We’re [Jerry](https://www.linkedin.com/in/jerry-n-zhang/) and [Cole](https://www.linkedin.com/in/colegawin/), co-founders of [Lemma](https://www.uselemma.ai/) (YC F25).\n\n**The Problem:**\n\n**AI agents don’t learn from their mistakes. In fact, they get worse with use.**\n\nIn production, prompts and agents continuously degrade due to real-world input drift (new user behaviors or unseen edge cases). Agent performance can often **drop \\~40% in a few weeks**, and suddenly what worked in testing breaks in front of customers.\n\nWhen that happens, engineers are forced to dig through logs, collect failing examples, and manually tweak prompts rather than building core product features.\n\n**Solution:**\n\nThat’s why we built **Lemma: the first end-to-end system that closes the loop between agent deployment and improvement.**\n\n**Here's what that means:**\n\n**Step 1:** Lemma detects failed outcomes directly from live traffic, and it automatically identifies the exact cause in an agent chain.\n\n**Step 2:** Lemma alerts you, and with one click, it runs targeted prompt optimizations to fix the failing behavior without any manual tracing or guesswork.\n\n**Step 3:** We give you back an improved prompt and automatically open a PR in your codebase so your prompts can live where you want them. Alternatively, you can also fetch your prompt from the Lemma dashboard.\n\n**Plus, Lemma provides all the LLM eval and observability features you rely on, just reimagined for continuous learning:**\n\n* Data-ingestion pipeline to bring your existing eval sets and automatically flag inconsistencies and gaps\n\n![uploaded image](/media/?type=post&id=95219&key=user_uploads/1798805/4f862c09-dce0-48da-91fe-1a152a5d82da)\n\n* Prompt editor and inference support for any closed & open-source model for prompt iteration.\n\n![uploaded image](/media/?type=post&id=95219&key=user_uploads/1798805/92464a28-ab54-4511-bc79-fd3e4f9d4823)\n\n* Agent tracing observability with live drift detection, regression alerts, and performance visibility across real user interactions.\n\n![uploaded image](/media/?type=post&id=95219&key=user_uploads/1798805/2b82ecee-1965-4fc0-831d-36b76db3f064)\n\nTeams using Lemma cut manual prompt iteration by 90%, resolve production drifts in minutes instead of days, and improve model performance \\~2–5% every optimization cycle.\n\n**Our Story:**\n\nWe met freshman year at USC and have been building together ever since instead of going to classes.\n\nBefore starting Lemma, we were engineers at two high growth, AI-native startups: Tandem (AI for healthcare) and Chipstack (AI agents for chip design). At both companies, setting up evaluations looked like clunky Retool dashboards and multiple engineers manually tweaking experiments. We built internal systems that automated both running the evaluations themselves, as well as the error-driven feedback loop. The result: 2x accuracy improvement and speed of iteration.\n\nWe soon realized every AI company was reinventing the same internal tooling in-house. So we left college, joined YC, and are now bringing continuous learning infrastructure to everyone else.\n\n![uploaded image](/media/?type=post&id=95219&key=user_uploads/1798805/53754495-215a-47aa-884d-bae4a59f9a68)\n\n**Ask**\n\n**Try our platform** - If you’re building with LLMs and run a ton of prompt or eval experiments, we’d love for you to work with us.\n\n**Introductions** - If you know a Head of AI/Eng or CTO at a pre-seed to Series A startup, we owe you lunch :)\n\nPlease reach out at [jerry@uselemma.ai](mailto:jerry@uselemma.ai) or book a live demo on our website [uselemma.ai](http://uselemma.ai). All help is appreciated - thank you!\n\n![uploaded image](/media/?type=post&id=95219&key=user_uploads/1798805/53c71689-276c-4993-b693-50aa8a492566)\n\n", "url": "https://www.ycombinator.com/launches/Oln-lemma-continuous-learning-for-ai-agents", "votes": 204, "created_at": "2025-11-05T19:58:08.940Z" } ] }, { "id": 31261, "name": "Control Seat", "slug": "control-seat", "former_names": [ "C42" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/c9fbf6b23216653c383d7f0caea2def1b673a3a5.png", "website": "https://www.controlseat.com", "all_locations": "San Francisco, CA, USA", "long_description": "Control Seat builds the hardware and software that helps industrial teams see, understand, and run their operations.\r\n\r\nOur platform creates one clear source of truth across the plant. It connects the systems teams already use, brings control, monitoring, history, and machine health together, and adds an intelligence layer that continuously watches for meaningful changes. Teams can understand what happened, catch problems earlier, and act before they become downtime.\r\n\r\nThe result is fewer surprises, faster decisions, and less time spent piecing together information across disconnected tools.\r\n\r\nWe work with manufacturers, utilities, energy companies, and other industrial operators modernizing existing facilities or building new ones.", "one_liner": "AI that runs factories. ", "team_size": 2, "industry": "Industrials", "subindustry": "Industrials", "launched_at": 1777448165, "tags": [ "Workflow Automation", "Monitoring", "Industrial" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "status": "Active", "industries": [ "Industrials" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/control-seat", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/control-seat.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/control-seat", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/c9fbf6b23216653c383d7f0caea2def1b673a3a5.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Jared Friedman", "url": "https://www.ycombinator.com/people/jared-friedman" }, "founders": [ { "user_id": 3064044, "full_name": "Warren Shepard", "title": "Founder/CEO", "founder_bio": "Building a modern SCADA system to control and monitor factory hardware and replace legacy systems. Previously at Dartmouth ('26), SpaceX (Starlink), and Microsoft (security).", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/warren-shepard-493a37201/", "x_url": "https://x.com/warrren_shepard", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ff41fa222f3fc3b27bd9e580e1dee71f5754722d.jpg" }, { "user_id": 3132492, "full_name": "Jack Grodnick", "title": "Founder/CTO", "founder_bio": "Building Control Seat, the AI-native controls and monitoring stack for industry. We help industrial teams connect plant systems, understand what changed, catch problems earlier, and run operations with fewer surprises. On leave from Dartmouth. Previously at Tesla, SpaceX, and Amazon.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/jack-grodnick/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b49172bc391f2b19dcde4aff68ff061ccd456a44.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 104972, "title": "Control Seat: the operating system for the physical world.", "tagline": "An AI-native controls (SCADA) platform and integration service.", "body": "\\\n\\\n**THE PROBLEM**\n\nIndustrial control software still looks and feels like it was built in the Windows 95 era. Manufacturers spend weeks stitching together Programable Logic Controllers (PLCs), Supervisory Control and Data Acquisition (SCADA), HMIs, databases, and custom integrations just to get systems running. Additionally, legacy systems collect enormous amounts of high-quality data, but don’t do _anything_ with it, even though it could easily be used for things such as anomaly detection or optimization recommendations.\\\n\\\n**OUR SOLUTION**\n\nWe're building an AI-native platform to unify the industrial control stack, starting with SCADA and eventually extending down to PLC hardware and the devices on the factory floor. We offer both a product and a service:\n\n**1) Software**\n\nWe already built a full SCADA system to replace legacy platforms like Ignition, WinCC, and FactoryTalk. It’s AI native: users can use AI to generate controls screens in seconds (this used to take hours), write custom scrips/logical flows, and automatically detect issues before they become catastrophic. \\\n\\\nNot only is this a full SCADA system, but it can operate on top of an existing one so that users can take advantage of our “intelligence layer” without having to rip out their current infrastructure (people understandably don’t like doing this).\\\n\\\n**2) System Integration** \n\nWe’ve recently expanded to become an AI-native industrial integrator. Instead of just selling our software, we integrate it into our clients existing stack for them, and help them set up any other tools they need to run it. \n\nOver the past few months, we've spent hundreds of hours talking with controls engineers, manufacturers, operators, and system integrators. We already knew that industrial software was broken, but we also learned that even when better technology exists, people are understandably cautious about adopting it. These systems run critical infrastructure and reliability matters more than novelty. Change is slow, but we want to reindustrialize America now.\n\n**Instead of waiting for the industry to adopt an entirely new software stack, we're bringing our technology directly into customer projects today.**\n\nWhile larger firms can spend millions on integration services, smaller firms don’t have that kind of money. Traditional integrators use highly manual methods to set up their clients software stacks (which are also outdated). By leveraging AI, we’ll be able to set up the same systems at 10x the speed and charge way less. \n\n**Our ask: if you’re exploring controls/automation systems, PLC programming, SCADA systems, or system integration**, **we’d love to talk. Or, if you know any deep tech companies that want to set up a controls system, let us know!**\n\nSee more at .", "url": "https://www.ycombinator.com/launches/RJ6-control-seat-the-operating-system-for-the-physical-world", "votes": 4, "created_at": "2026-07-09T14:31:33.522Z" } ] }, { "id": 33019, "name": "Agnost AI", "slug": "agnost-ai", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/e1595f92ea75b1f367addedb5703852b1c7efded.png", "website": "https://agnost.ai", "all_locations": "San Francisco, CA, USA", "long_description": "Agnost AI reads every conversation a company's AI Agent has with its users to find where it's silently failing and then turns that same data into custom models that run the agent better, faster & cheaper than frontier models.", "one_liner": "We find where AI agents fail, then train better models to run them.", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Analytics", "launched_at": 1783481375, "tags": [ "Developer Tools", "Machine Learning", "Analytics", "Monitoring", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "status": "Active", "industries": [ "B2B", "Analytics" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/agnost-ai", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/agnost-ai.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://linkedin.com/company/agnostai", "x_url": "https://x.com/agnostai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/e1595f92ea75b1f367addedb5703852b1c7efded.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/b09996d90ed38a4b5d62581651fffd7e0885e6b2.png" ], "primary_partner": { "name": "Tyler Bosmeny", "url": "https://www.ycombinator.com/people/tyler-bosmeny" }, "founders": [ { "user_id": 1987056, "full_name": "Shubham Palriwala", "title": "Founder/CEO", "founder_bio": "Co-founder and CEO of Agnost AI.\nYoungest Engineer on Cisco's Analytics team, First Hire at Formbricks (Open-Source Qualtrics), wrote code in Bitcoin & other open-source projects across OWASP and the Linux foundation.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/shubhampalriwala/", "x_url": "https://x.com/ShubhamInTech", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/16f7d9652057acfedcb3ad0c94ecab3e15758aba.jpg" }, { "user_id": 2055517, "full_name": "Parth Ajmera", "title": "Founder/CTO", "founder_bio": "Co-Founder/CTO at Agnost AI. Graduated in Computer Science from IIT Madras, ranking 159 among over 1 million candidates. Previously led graphics engineering at Infurnia and built terabyte scaled data pipelines at Microsoft", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/parthajmera/", "x_url": "https://x.com/prrthh132", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/93dc08b7da172c3288f1f798f2b9a0b27db75f16.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 111310, "title": "Agnost AI: Turn Your Agent Traces Into a Faster, Cheaper, & more Accurate Custom Model", "tagline": "We use your agent’s production traces to train a specialized model that performs better, responds faster, and costs less.", "body": "Your agent logs aren’t just for debugging. They’re your training data.\\\nHey YC! We’re Shubham and Parth, childhood friends and founders of Agnost AI.\n\n**TL;DR**: Agnost AI uses your agent’s production traces to train a model specialized for your exact workload. Our first custom model for our customer delivered them a:\n\n* 22.9% higher task success,\n* 90.2% lower median latency, and\n* 94.5% lower cost than Opus 4.8.\n\nDemo: \n\n### What we found\n\nA few weeks ago, we launched Agnost AI as product analytics for AI agents. We read conversations between users and agents to uncover failures, behavior drift, hallucinated links, frustration, and churn signals.\n\nWe expected to find bugs. Surprisingly, we also found a lot of infrastructure waste. Many customer-facing agents use frontier models for narrow, repetitive jobs:\n\n* Route a request to the correct tool\n* Have a narrow system prompt with the given inputs\n* Retrieve the relevant data from internal tool calls\n* Parse & format the final response\n\nThe underlying frontier model can write code, solve advanced mathematics, understand biology, and reason across hundreds of domains. **Your agent uses a small, repeatable slice of that capability while paying frontier-model prices for every request.**\n\n### We trained a specialist\n\nWe used historical production traces to train our first workload-specific model for an early customer: **agnost-<redacted>-0.1**\n\nWe evaluated it against Opus 4.8 on 780 held-out customer traces for an Identifier Extraction Agent. The results:\n\n* Task success: 71.5% → 87.9%, 22.9% relative improvement\n* Median latency: 4.80s → 0.47s, 90.2% lower\n* Tail latency: 11.20s → 1.35s, 87.9% lower\n* Cost per 1,000 completed tasks: $42.00 → $2.30, 94.5% lower\n\nThis is one model for one workload, not a claim that specialist models should replace frontier models everywhere.\n\nFrontier models remain the right choice for open-ended work. But once an agent’s production job becomes predictable, **paying for general intelligence on every request stops making sense**.\n\n### How it works\n\n1. Send your existing traces to Agnost through OpenTelemetry or our SDK.\n2. We identify the recurring workflows your agent performs.\n3. We create an eval set from held-out production traces.\n4. We train a specialist model on the historical workload.\n5. We benchmark it against your current model on success, latency, and cost.\n\nYou only consider switching when the specialist wins on your own workload. Your logs go from debugging data to evaluation data and finally to training data recursively.\n\n### Who we are\n\nWe’re childhood friends, eight years and counting.\n\n* Shubham: Youngest engineer on Cisco’s analytics team, first hire at Formbricks, and contributor to Bitcoin, OWASP, and the Linux Foundation.\n* Parth: IIT Madras CS, ranked 159 out of 1.5 million. Built Spark pipelines at Microsoft and led GPU technology at Infurnia.\n\n### Our ask\n\nIf you have meaningful daily traffic, send us an anonymized trace export. We’ll analyze the workload, build a held-out evaluation set, and tell you whether a specialist model is worth training. If it is, we’ll benchmark it against your current frontier model.\n\nEmail me at [shubham@agnost.ai](mailto:shubham@agnost.ai) or book a call at [call.agnost.ai](http://call.agnost.ai)\n\nMore about us at: [agnost.ai](https://agnost.ai)\n\n![uploaded image](/media/?type=post&id=111310&key=user_uploads/1987056/6599ed29-ddf9-4d02-9232-425af1535be7)\n\n", "url": "https://www.ycombinator.com/launches/SxK-agnost-ai-turn-your-agent-traces-into-a-faster-cheaper-more-accurate-custom-model", "votes": 97, "created_at": "2026-08-20T16:00:00.252Z" }, { "id": 107260, "title": "Agnost AI: Product Analytics for AI Agents", "tagline": "We read every conversation your users have with your agents & chatbots and surface hidden insights & product feedback. We are working with teams at Google, Exa and Corgi", "body": "**Your users are telling your agent exactly why they churn! You’re not listening.**\n\nHey YC! We’re Shubham and Parth, childhood friends, founders of **Agnost AI**.\n\n**TL;DR: Agnost AI reads every conversation your chat or voice agent has with your users. We show you what your users want, where they get frustrated, and why they leave.**\n\n**The feedback hiding in your logs becomes discoveries, evals, and fixes ready to ship. We’re already processing over 1 million events every day and growing.**\n\nTry us out at[ agnost.ai](http://agnost.ai)\n\n### \n\n### **The problem**\n\nTraditional product analytics were built for clicks.\\\nObservability was built for token costs, latencies, and deterministic errors.\n\nBut every product is rapidly moving from UI to conversation, and agent failures don’t look like traditional failures. The LLM returns 200 OK and says “sorry, I can’t do that.” The real failures are hiding in your conversations:\n\n* “can it remember this for next time?” → feature request\n* “you said it’s done, but nothing happened” → bug\n* “i’ve told you this three times already” → frustration\n* “nvm i’ll just do it myself” → churn\n\nToday teams throw more evals at this. But evals only tell you whether you fixed something, not what to fix. You can’t write a test for a problem you haven’t discovered yet.\n\n### **How it works**\n\n1. **Install our skill:** 3 lines of code, or point your OpenTelemetry exporter at us. Any LLM, any framework. Add any custom metadata like user tier, model versions, etc. We’re at:[ docs.agnost.ai](http://docs.agnost.ai)\n2. **Every conversation becomes a signal:** feature requests, bugs, frustration, churn risk, failed upgrades. Detected for your product, with the exact users & conversations behind each.\n3. **The insight becomes action:** when hundreds of users rageprompt at your agent due to a specific bug, Agnost AI hands you the insight with the exact conversations behind it: \n * create an eval from it\n * debug it,\n * or let Agnost fix it itself: our GitHub app opens a PR, tested against your past conversations, so you know it works before it ships.\n\n Your agent gets better every week.\n\n### **Who we are**\n\nChildhood friends, 8 years and counting.\n\n![uploaded image](/media/?type=post&id=107260&key=user_uploads/2055517/e3c2eaad-8b6e-4ec0-abfa-b16074688963)\n\n**Shubham**: Youngest engineer on Cisco’s analytics team, First Hire at Formbricks, Wrote code in Bitcoin, OWASP, and the Linux Foundation.\n\n**Parth**: IIT Madras CS (ranked 159 of 1.5M), built Spark pipelines at Microsoft, Led GPU tech at Infurnia (\\~$1M ARR).\n\n### **Our Asks**\n\n1. **Building a chat or voice agent?** Sign up & I’ll personally find 3 insane things going on in your agent traces that you’d have no idea about.\n2. **Know a founder whose product is an agent?** Intro us, 40’ monitors (to monitor your agents) on us: [shubham@agnost.ai](mailto:shubham@agnost.ai)\n3. **Try it:** Free tier:[ agnost.ai](http://agnost.ai)", "url": "https://www.ycombinator.com/launches/Ru0-agnost-ai-product-analytics-for-ai-agents", "votes": 191, "created_at": "2026-07-24T15:53:30.811Z" } ] } ]