[ { "id": 11948, "name": "Centaur", "slug": "centaur", "former_names": [ "Centaur Labs" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/a8bf2233f3d639f46c87be05e11eb153ca925d88.png", "website": "https://centaur.ai/", "all_locations": "Boston, MA, USA", "long_description": "The best AI models aren’t just trained and evaluated with human data; they’re built with superhuman data. The strongest datasets emerge through collective intelligence, where humans and machines work together to outperform either one alone. At Centaur, we create superior quality data by turning annotation into an arena where experts and AI compete.", "one_liner": "We Create Superior Data By Making Annotation Competitive", "team_size": 45, "industry": "Healthcare", "subindustry": "Healthcare -> Healthcare IT", "launched_at": 1541050478, "tags": [ "Data Science", "Crowdsourcing", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2019", "status": "Active", "industries": [ "Healthcare", "Healthcare IT" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Growth", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/centaur", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2019/centaur.json", "year_founded": 2017, "city": "Boston", "country": "US", "linkedin_url": "https://www.linkedin.com/company/centaur-labs/", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/a8bf2233f3d639f46c87be05e11eb153ca925d88.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 292290, "full_name": "Erik Duhaime", "title": "Founder", "founder_bio": "founder/CEO at Centaur.AI", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/erik-duhaime-7a827413/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/9a6e50d5dd71f76d5a040c9cf9446f6cb36cdec8.jpg" }, { "user_id": 306580, "full_name": "Tom Gellatly", "title": "Founder", "founder_bio": "Hello! I live in Durham, NC. I have a background in mobile and web development and engineering management. I studied CSE at the University of Michigan. I was the first employee at SF ride-sharing startup Sidecar, which was acquired by General Motors in 2016. I then managed the mapping & data labeling engineering teams at Cruise Automation for 1.5 years.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/tomgellatly", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/a8c82270623bde9acd62fce72fb514ed762a7688.jpg" }, { "user_id": 283341, "full_name": "Zach Rausnitz", "title": "Founder", "founder_bio": "I'm a self-taught, generalist software engineer. I cofounded Centaur Labs in 2017. Previously I worked as a full-stack software engineer at Washington City Paper, where many notable people in media got their start.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/zach-rausnitz-55086927/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/5d9d3ff5deaaf87ce5c4bbdbe2e9858a6259919f.jpg" } ], "jobs": [ { "id": 114162, "title": "Technical Product Manager, Labeling Tools", "url": "https://www.ycombinator.com/companies/centaur/jobs/8ElCTgv-technical-product-manager-labeling-tools", "location": "US / Remote (US)", "type": "Full-time", "role": "Product", "role_type": null, "salary_range": "$145K - $160K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Gamifying medical data labeling to advance AI | MIT News | Massachusetts Institute of Technology", "url": "https://news.mit.edu/2023/gamifying-medical-data-labeling-ai-0628", "date": "Jun 28, 2023" }, { "title": "Centaur Labs Gets $15 Million To Improve Data For Healthcare AI", "url": "https://www.forbes.com/sites/gilpress/2021/09/03/centaur-labs-gets-15-million-to-improve-data-for-healthcare-ai/?sh=766f060063a0", "date": "Sep 06, 2021" } ], "launches": [] }, { "id": 13372, "name": "Anatomic", "slug": "anatomic", "former_names": [ "Localpay - Bizrise Technologies Pvt Ltd", "NearPe - Bizrise Technologies Pvt Ltd", "NearPe", "NearPe", "NearPe Payments Inc", "NearPe Payments", "Bikry" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/933a9327631b36a120754ee435d9763089cc6d8a.png", "website": "https://askanatomic.com", "all_locations": "Gurugram, HR, India; Noida, UP, India", "long_description": "Anatomic works with hospitals and doctors to collect surgical video, doctor-patient conversations and medical imaging for AI and robotics research. Our data library includes multi-perspective surgical recordings, diarized and transcribed clinical conversations, and annotated X-ray, CT and MRI series linked to longitudinal patient records. Built through a network of 100+ hospitals and 1,000+ clinicians, with HIPAA- and GDPR-compliant data.", "one_liner": "Real-world healthcare data for medical AI", "team_size": 3, "industry": "Healthcare", "subindustry": "Healthcare -> Healthcare IT", "launched_at": 1583756896, "tags": [ "B2B", "Digital Health", "Healthcare", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2020", "status": "Active", "industries": [ "Healthcare", "Healthcare IT" ], "regions": [ "India", "South Asia" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/anatomic", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2020/anatomic.json", "year_founded": 2020, "city": "Gurugram", "country": "IN", "linkedin_url": "https://www.linkedin.com/company/askanatomic", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/933a9327631b36a120754ee435d9763089cc6d8a.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 138434, "full_name": "Abhishek Bhayana", "title": "Founder/CEO", "founder_bio": "Founder & CEO of Anatomic, which provides real-world healthcare data (surgical video, clinical conversations and medical imaging) for medical AI. Previously founded Bikry and StrawberryPOS, and worked in private equity in New York. Purdue alum.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/abhayana", "x_url": "https://twitter.com/AbhishekBhayana", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/40e7c55ba3ba51f12d73ff7413211e0e474e3c28.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 24879, "name": "Lightly", "slug": "lightly", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/5c14e2db32ad93491460c405faf7230c30575e54.png", "website": "https://lightly.ai/", "all_locations": "Zürich, ZH, Switzerland", "long_description": "When ML teams send their data to companies like Scale.ai for labeling, most can only afford to label 1% or less of their datasets. But today they don’t have a good way to pick which 1% to label. We help them pick the best 1% of their data to label. By labeling the most representative data, they significantly improve model accuracy at the same cost.\r\n", "one_liner": "Help ML teams label the right data", "team_size": 5, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1629386219, "tags": [ "Machine Learning", "Data Labeling" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Summer 2021", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "Switzerland", "Europe", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/lightly", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2021/lightly.json", "year_founded": 2018, "city": "Zürich", "country": "CH", "linkedin_url": "https://www.linkedin.com/company/lightly-tech", "x_url": "https://twitter.com/LightlyAI", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/mirage", "github_url": "https://github.com/lightly-ai", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/5c14e2db32ad93491460c405faf7230c30575e54.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Brad Flora", "url": "https://www.ycombinator.com/people/brad-flora" }, "founders": [ { "user_id": 332413, "full_name": "Igor Susmelj", "title": "Founder", "founder_bio": "Igor has more than five years of experience in machine learning. He holds a degree in electrical engineering from ETH Zurich. During his studies, he developed a lot of experience in machine learning and robotics and had multiple successful publications in the area of deep learning at top conferences such as ICML and ECCV. He previously worked for two years at the Swiss stock exchange as a software engineer.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/igorsusmelj/", "x_url": "https://twitter.com/ISusmelj", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/1d9349e0fc104753bb1766db702c87fd90497a13.jpg" } ], "jobs": [ { "id": 48992, "title": "Account Executive in a fast-growing AI Startup", "url": "https://www.ycombinator.com/companies/lightly/jobs/U2seZC7-account-executive-in-a-fast-growing-ai-startup", "location": "San Francisco / Zürich, Zurich, CH / London, England, GB / Remote (GB; CA; US; CH)", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$150K - $300K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 84374, "title": "Business Development Representative in a fast-growing AI Startup", "url": "https://www.ycombinator.com/companies/lightly/jobs/64Uajzb-business-development-representative-in-a-fast-growing-ai-startup", "location": "GB / US / CH / Remote (GB; US; CH)", "type": "Full-time", "role": "Support", "role_type": null, "salary_range": "$80K - $300K", "equity_range": "0.15% - 0.35%", "experience": "Any (new grads ok)", "visa": "US citizenship/visa not required", "skills": [] } ], "news": [ { "title": "Lightly Promises Faster Food For AI Models After $3 Million Seed Round", "url": "https://www.forbes.com/sites/davidprosser/2022/05/11/lightly-promises-faster-food-for-ai-models-after-3-million-seed-round/?sh=791da54c4626", "date": "May 11, 2022" } ], "launches": [] }, { "id": 25307, "name": "Sieve", "slug": "sieve", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/b247c36fd7cd016d5bb505d32429cd07b9acfcbb.png", "website": "https://sievedata.com/", "all_locations": "San Francisco, CA, USA", "long_description": "Sieve builds the data and environments frontier AI labs use to train the next generation of multimodal systems.\r\n\r\nAI is moving beyond chatbots into video, audio, images, software, robotics, and interactive worlds. The next generation of models will need to understand how the world looks, sounds, moves, responds, and changes over time. Progress is bottlenecked by one thing: high-quality data.\r\n\r\nSieve brings together exabyte-scale infrastructure, novel multimodal understanding techniques, large-scale sourcing, and deep research partnerships to create datasets and environments with unmatched precision, quality, and speed. This has earned the trust of frontier AI labs, Fortune 100 companies, and fast-growing AI startups working on generative media, robotics, computer use, world models, and agentic systems.", "one_liner": "The multimodal data lab", "team_size": 34, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1643510053, "tags": [ "Developer Tools", "Video", "Data Labeling", "Data Engineering", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2022", "status": "Active", "industries": [ "B2B", "Infrastructure" ], "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/sieve", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2022/sieve.json", "year_founded": 2022, "city": "San Francisco", "country": "US", "linkedin_url": "https://linkedin.com/company/sievedata", "x_url": "https://twitter.com/@sievedata", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/b247c36fd7cd016d5bb505d32429cd07b9acfcbb.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": 412206, "full_name": "Mokshith Voodarla", "title": "Founder", "founder_bio": "Mokshith is co-founder and CEO of Sieve. Mokshith used to work on computer-vision problems at Scale AI, NVIDIA, and Ford where he experienced first-hand, the difficulties of building and scaling vision-based analytics systems. Before that, he graduated from UC Berkeley with a B.S. Electrical Engineering and Computer Science. Mokshith found his love for computer-vision on his high school robotics team, where he also met Abhi. Mokshith deeply enjoys orange chicken.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/mokshith-voodarla/", "x_url": "https://twitter.com/mokshith_v", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ed90fadcb7d40c3b39a55d49c9049f636e06fa29.jpg" }, { "user_id": 858478, "full_name": "Abhinav Ayalur", "title": "Founder", "founder_bio": "Abhinav (Abhi) Ayalur is the co-founder and CTO of Sieve. Abhi leads technical design and development along with steering the overall company vision. Abhi's previous experience includes stints in computer vision, data analysis, and API development at Second Spectrum, NVIDIA, Microsoft, and Niantic. Abhi graduated UC Berkeley with a B.S. in Electrical Engineering and Computer Science. In his spare time, Abhi loves to play the alto saxophone, eat good food, and play basketball.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/abhi12-ayalur/", "x_url": "https://twitter.com/thpicy", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ae4830f3e54dadcd5d67c2f4045e3732580975cc.jpg" } ], "jobs": [ { "id": 75776, "title": "Distributed Systems Engineer", "url": "https://www.ycombinator.com/companies/sieve/jobs/6UaDNRn-distributed-systems-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $300K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 75777, "title": "Applied Research Engineer", "url": "https://www.ycombinator.com/companies/sieve/jobs/epe3f1w-applied-research-engineer", "location": "San Francisco", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $300K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 81029, "title": "Software Engineer", "url": "https://www.ycombinator.com/companies/sieve/jobs/8viV0w4-software-engineer", "location": "San Francisco", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $300K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 82322, "title": "Product & Ops Lead ", "url": "https://www.ycombinator.com/companies/sieve/jobs/QO2KXMO-product-ops-lead", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$130K - $250K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 89312, "title": "Product Engineer", "url": "https://www.ycombinator.com/companies/sieve/jobs/R7esADT-product-engineer", "location": "San Francisco", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $300K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 89313, "title": "Recruiting", "url": "https://www.ycombinator.com/companies/sieve/jobs/4vb50L5-recruiting", "location": "San Francisco", "type": "Full-time", "role": "Recruiting & HR", "role_type": "Recruiter", "salary_range": "$120K - $200K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 89736, "title": "Reliability Engineer", "url": "https://www.ycombinator.com/companies/sieve/jobs/2mp3sT5-reliability-engineer", "location": "San Francisco", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$150K - $300K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 91896, "title": "Data Operations Lead", "url": "https://www.ycombinator.com/companies/sieve/jobs/gpArec4-data-operations-lead", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$130K - $250K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Sieve's Video AI API Beta and ~$4M Raise | Sieve Blog - Video AI | Sieve", "url": "https://www.sievedata.com/blog/launch", "date": "Nov 14, 2022" } ], "launches": [] }, { "id": 25433, "name": "JumpWire", "slug": "jumpwire", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/077148a8281f282d5a5d6825d93aedf0b71c08b4.png", "website": "https://jumpwire.io", "all_locations": "New York City, NY, USA; New York, NY, USA; Remote", "long_description": "JumpWire is a data protection platform that adds advanced data security controls between APIs, applications and databases. JumpWire automatically identifies sensitive properties inside large data sets and gives developers full control over which people and applications can access or update records containing sensitive info.\r\n\r\nExamples uses include restricting who can read customer PII to members of the customer service team, giving on-call engineers elevated access to production, or splitting user records between regions for GDPR purposes.\r\n\r\nJumpWire’s approach to securing data in-place minimizes the risk of data leaks exposing sensitive information or mishandling by other applications and vendors. The exact security scheme applied to data is defined by policies that align with an organization’s existing InfoSec program.\r\n\r\nJumpWire helps companies who maintain information security with compliance programs such as SOC or HIPAA. They are processing sensitive data, often from their own customers, and exceed security best practices as a competitive advantage. JumpWire provides defense at depth to data and sits alongside access controls and Layer 4 encryption to provide a comprehensive data security solution.\r\n\r\nJumpWire is unique from solutions such as data vaults by installing inside our customers’ own infrastructure and clouds. It is interoperable with existing applications and databases, which eliminates the need for large data migrations or code refactoring. Lower-level approaches to data security, such as encryption at rest, are too blunt and lack the ability to differentiate between properties in the data itself. Its scope is limited to physical storage, and security is lost as soon as an application or query loads the data.", "one_liner": "Dynamic access controls for all data and databases", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Security", "launched_at": 1644446688, "tags": [ "Security", "Data Labeling", "Databases" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2022", "status": "Acquired", "industries": [ "B2B", "Security" ], "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/jumpwire", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2022/jumpwire.json", "year_founded": 2021, "city": "New York City", "country": "US", "linkedin_url": "https://www.linkedin.com/company/jumpwire", "x_url": "https://twitter.com/jumpwire_ai", "fb_url": null, "cb_url": null, "github_url": "https://github.com/extragoodlabs", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/077148a8281f282d5a5d6825d93aedf0b71c08b4.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 823026, "full_name": "Ryan Cooke", "title": "Founder", "founder_bio": "Ryan has spent his career working in startups, with a passion for creating innovative technology products. JumpWire participated in the YC W22 batched, and was acquired by WorkOS in 2024. Ryan has previously co-founded and led engineering teams at Docurated, N26 and Orum (acquired by Stripe). He's obsessed with running his own AI, and particularly enjoys building teams aside software.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/ryancooke/", "x_url": "https://x.com/mapandfold", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/36cbdb63975aa38768f2c1241ebb5034d7c15861.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 25948, "name": "Spade", "slug": "spade", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/dc25342edef4b9f6942587780f3a5d96fae04f9c.png", "website": "https://www.spade.com", "all_locations": "New York City, NY, USA", "long_description": "We take messy transaction data and turn it into structured, verified records — with a platform, models, and tooling that help our customers use it everywhere it matters.", "one_liner": "Spade is the data and AI platform for modern finance.", "team_size": 30, "industry": "Fintech", "subindustry": "Fintech", "launched_at": 1641512856, "tags": [ "Fintech", "Machine Learning", "Payments", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2022", "status": "Active", "industries": [ "Fintech" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Growth", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/spade", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2022/spade.json", "year_founded": 2021, "city": "New York City", "country": "US", "linkedin_url": "https://www.linkedin.com/company/spadedata/", "x_url": "https://twitter.com/Spadedata", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/spade", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/dc25342edef4b9f6942587780f3a5d96fae04f9c.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": 794205, "full_name": "Oban MacTavish", "title": "Founder", "founder_bio": "Oban MacTavish is the CEO of Spade. Before starting Spade Oban co-founded Hubly - a vertical SaaS platform for independent RIAs. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/oban/", "x_url": "https://twitter.com/Obanmact", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/865ea81c5261591da4c1d4c61a26fae1a2014dce.jpg" } ], "jobs": [ { "id": 98588, "title": "Backend Engineer (Mid/Sr)", "url": "https://www.ycombinator.com/companies/spade/jobs/k9ZLlCc-backend-engineer-mid-sr", "location": "New York, NY", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$170K - $230K", "equity_range": "", "experience": "5+ years", "visa": "US citizen/visa only", "skills": [ "Python" ] }, { "id": 98589, "title": "Finance Lead", "url": "https://www.ycombinator.com/companies/spade/jobs/ahCOFcZ-finance-lead", "location": "New York, NY / Remote (US)", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$170K - $200K", "equity_range": "", "experience": "5+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 98592, "title": "Technical Customer Support", "url": "https://www.ycombinator.com/companies/spade/jobs/2rT8fjC-technical-customer-support", "location": "New York, NY / Remote (US)", "type": "Full-time", "role": "Support", "role_type": null, "salary_range": "$85K - $108K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Spade raises $40M for transaction data enrichment", "url": "https://www.axios.com/pro/fintech-deals/2026/03/24/spade-40m-bank-transaction-data-tools", "date": "Mar 24, 2026" }, { "title": "Fintech Spade's Series a Pitch Deck Used to Raise $10 Million", "url": "https://www.businessinsider.com/fintech-spades-series-a-pitch-deck-raise-10-million-2023-12", "date": "Dec 18, 2023" }, { "title": "Spade digs into credit card fraud detection intelligence following new capital raise | TechCrunch", "url": "https://techcrunch.com/2023/12/06/spade-credit-card-fraud-detection/", "date": "Dec 06, 2023" }, { "title": "Spade raises $5m in seed funding round led by a16z", "url": "https://blog.spade.com/spade-raises-5m-in-seed-funding-led-by-a16z/", "date": "Feb 27, 2023" }, { "title": "Spade turns credit card transaction gibberish into clear, actionable data | TechCrunch", "url": "https://techcrunch.com/2023/02/27/spade-credit-card-data-a16z/", "date": "Feb 27, 2023" } ], "launches": [] }, { "id": 28787, "name": "Deasy Labs", "slug": "deasy-labs", "former_names": [ "Deasie" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/b2f7a8dffc7dc8a6fdb4c13dce765cc7d2820b9a.png", "website": "https://www.deasylabs.com/", "all_locations": "New York City, NY, USA", "long_description": "Deasy Labs was acquired by Collibra in July 2025 (global leader in enterprise data governance).\r\n\r\nDeasy Labs provides metadata orchestration for AI workflows. Deasie's platform provides the best way for AI teams to create and embed high-quality, customized metadata into their AI workflows (e.g., RAG, Agentic frameworks).\r\n\r\nOur three founders (from Amazon, McKinsey/QuantumBlack & MIT) previously built an ML data governance tool from 0 to 1 within McKinsey, which we deployed with 11 Fortune 500 companies. We saw in early 2023 the ability to create high-quality metadata (without reliance on domain experts) would be a key factor in achieving the accuracy & speed in GenAI applications required for production. \r\n\r\nOur investors include General Catalyst, Y Combinator, RTP Global and world experts in enterprise data. Website: https://deasylabs.com", "one_liner": "Metadata for GenAI workflows", "team_size": 8, "industry": "B2B", "subindustry": "B2B -> Analytics", "launched_at": 1691123582, "tags": [ "Data Labeling", "Big Data", "AI", "AI Assistant", "Databases" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2023", "status": "Acquired", "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/deasy-labs", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2023/deasy-labs.json", "year_founded": 2023, "city": "New York City", "country": "US", "linkedin_url": "https://www.linkedin.com/company/deasie", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/b2f7a8dffc7dc8a6fdb4c13dce765cc7d2820b9a.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": 1227239, "full_name": "Reece Griffiths", "title": "Founder", "founder_bio": "Co-founder of Deasy Labs (YC S23): Data governance for language models in the enterprise. \n\nExited Deasy Labs to Collibra in < 2 yrs (global leader in data governance).\n\nMy documented startup journey: https://reecegriffiths.substack.com/\n\nPrior experience: QuantumBlack: Product Manager for AI for Data Governance product (1.5 yrs); McKinsey & Company: Management consultant (2 yrs); X26: Founder of X26 (global entrepreneurs community); University of Cambridge: Masters in information engineering.", "is_active": true, "linkedin_url": "https://linkedin.com/in/reece-william-griffiths", "x_url": "https://twitter.com/ReeceWGriffiths", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/53fe8f42805f99012d0a74523cf2874504663377.jpg" }, { "user_id": 1237667, "full_name": "Mikko Peiponen", "title": "Founder", "founder_bio": "Mikko is the co-founder of Deasie. He is a technologist with deep experience (5+ years) in data and machine learning space from McKinsey and QuantumBlack, where he implemented and deployed models at large enterprises. Prior to moving into the world of data governance for language models with Deasie, he was the Lead data scientist building AI data quality tool at QuantumBlack. Mikko holds a graduate degree from MIT.", "is_active": true, "linkedin_url": "https://linkedin.com/in/mikkopeiponen", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/9dd45b70aec24740409919d85c62d51a6b55da39.jpg" }, { "user_id": 1226121, "full_name": "Leonard Platzer", "title": "Founder", "founder_bio": "Co-founder of Deasy Labs: The Metadata Engine for search usecases. \n\nPrior experience: QuantumBlack: Software Engineer, Machine Learning at AI for Data Quality product; Previous Engineer at Amazon, Mercedes-Benz and ChatBot Startup (E-Bot 7); Bachelor from National University of Singapore and Graduate from UWC Mostar", "is_active": true, "linkedin_url": "https://linkedin.com/in/leonardplatzer", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/91e9bdf6844bdc43fac38dbfaa0bf7fedda307c0.jpg" } ], "jobs": [], "news": [ { "title": "Data governance platform Deasie raises $2.9 million in seed funding - Business Leader News", "url": "https://www.businessleader.co.uk/deasie-raises-2-9-million-seed-funding/", "date": "Oct 13, 2023" }, { "title": "Lakera and Deasie raise funding to make AI large language models more reliable - SiliconANGLE", "url": "https://siliconangle.com/2023/10/12/lakera-deasie-raise-funding-make-large-language-models-reliable/", "date": "Oct 12, 2023" }, { "title": "Deasie raises $2.9M for data governance platform expansion - Tech.eu", "url": "https://tech.eu/2023/10/12/deasie-raises-29-million-for-data-governance-platform-expansion/", "date": "Oct 12, 2023" }, { "title": "Deasie raises $2.9M for data governance platform expansion - Tech.eu", "url": "https://tech.eu/2023/10/12/deasie-raises-29-million-for-data-governance-platform-expansion/", "date": "Oct 12, 2023" }, { "title": "Deasie wants to rank and filter data to make generative AI more reliable | TechCrunch", "url": "https://techcrunch.com/2023/10/12/deasie-wants-to-rank-and-filter-data-to-make-generative-ai-more-reliable/", "date": "Oct 12, 2023" } ], "launches": [ { "id": 73808, "title": "Deasie - Data governance for language model applications", "tagline": "Is your enterprise data ready for language models?", "body": "**tl;dr: Deasie ensures that only relevant, high-quality, and safe data is fed into language models.**\n\nHi! We’re Reece, Leo, and Mikko and we’re excited to launch [Deasie](https://www.deasie.com).\n\n⚠️ **Problem**\n\nFor the first time, companies are turning towards their masses of unstructured data (e.g., documents, reports, emails) for a range of Generative AI use cases. Today, most companies are unable to answer the critical questions needed to ensure language models are trained and deployed reliably: Does this data contain sensitive information? Is this the most relevant data for this problem? Are there inconsistencies in the data that could skew my results?\n\n💡 **Solution**\n\nDeasie is a platform that provides automated checks for compliance (e.g., PII & proprietary data) and quality (e.g., irrelevant, untimely, or inconsistent information) of unstructured data, to reliably govern which data is used for which language model use case.\n\n👨‍👦‍👦 **The Team**\n\nFounders Leo, Mikko & Reece previously built McKinsey's award-winning product that used AI to enhance data quality in the enterprise, built ML applications across QuantumBlack, MIT, Amazon, and Mercedes, and have now set out to unlock the power of unstructured data for the upcoming wave on language model applications.\n\n![uploaded image](/media/?type=post&id=73808&key=user_uploads/1237667/ede65ff1-1431-4c6f-88aa-dd343d325f4d)\n\n👋 **Ask**\n\n**Lots of companies must be delving into the world of LLMs. If you’ve encountered challenges identifying and using high-quality input data (e.g., data that is relevant, timely, accurate, and consistent), then we’d love to hear from you! Please reach out to us at** [**leonard@deasie.com**](mailto:leonard@deasie.com)", "url": "https://www.ycombinator.com/launches/JCS-deasie-data-governance-for-language-model-applications", "votes": 16, "created_at": "2023-08-04T06:57:33.350Z" } ] }, { "id": 29659, "name": "Unbound", "slug": "unbound", "former_names": [ "Unbound Security" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/36df07c499b1ae5e15e4171c4cc3ef24d462d03c.png", "website": "https://getunbound.ai/", "all_locations": "San Francisco, CA, USA", "long_description": "", "one_liner": "Use AI tools without fear of data leakage", "team_size": 7, "industry": "B2B", "subindustry": "B2B -> Security", "launched_at": 1715812278, "tags": [ "Privacy", "Cybersecurity", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Summer 2024", "status": "Active", "industries": [ "B2B", "Security" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/unbound", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2024/unbound.json", "year_founded": 2023, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/91353269/", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/36df07c499b1ae5e15e4171c4cc3ef24d462d03c.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 901694, "full_name": "Rajaram Srinivasan", "title": "Founder/CEO", "founder_bio": "Building data security for Gen AI.\nPrev: Palo Alto Networks, Adobe | MIT Alum", "is_active": true, "linkedin_url": "https://linkedin.com/in/rajaramsrinivasan", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/a03990b831f31057f44e921c63774348d6c10eb1.jpg" }, { "user_id": 1985851, "full_name": "Vignesh Subbiah", "title": "Founder/CTO", "founder_bio": "Cofounder at Unbound. Security for AI agents.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/vigneshsubbiah16/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/390bb26fdd2b8135d1b2e6de0c675b9497c5ee6b.jpg" } ], "jobs": [ { "id": 69774, "title": "Founding Engineer", "url": "https://www.ycombinator.com/companies/unbound/jobs/FfwAxvH-founding-engineer", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Backend", "salary_range": "$110K - $180K", "equity_range": "0.10% - 1.00%", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 76243, "title": "Founding growth engineer", "url": "https://www.ycombinator.com/companies/unbound/jobs/YIPCJgp-founding-growth-engineer", "location": "San Francisco, CA, US / Remote (US)", "type": "Full-time", "role": "Sales", "role_type": null, "salary_range": "$80K - $135K", "equity_range": "0.10% - 0.50%", "experience": "1+ years", "visa": "US citizenship/visa not required", "skills": [] }, { "id": 93184, "title": "Software Engineering Intern", "url": "https://www.ycombinator.com/companies/unbound/jobs/hn66xKN-software-engineering-intern", "location": "San Francisco, CA, US", "type": "Internship", "role": "Engineering", "role_type": "Backend", "salary_range": "$5K - $7K / monthly", "equity_range": "", "experience": null, "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Unbound raises $4M to help enterprises embrace AI tools on their terms | Markets Insider", "url": "https://markets.businessinsider.com/news/stocks/unbound-raises-4m-to-help-enterprises-embrace-ai-tools-on-their-terms-1034775030", "date": "Jun 30, 2025" }, { "title": "Unbound secures USD $4 million seed to boost safe AI adoption", "url": "https://securitybrief.com.au/story/unbound-secures-usd-4-million-seed-to-boost-safe-ai-adoption", "date": "May 30, 2025" }, { "title": "Race Capital Leads USD 4 Mn Funding Round in Unbound | Entrepreneur", "url": "https://www.entrepreneur.com/en-in/news-and-trends/race-capital-leads-usd-4-mn-funding-round-in-unbound/492477", "date": "May 29, 2025" }, { "title": "Governance Start-Ups Boom In The Battle To Keep AI Honest", "url": "https://www.forbes.com/sites/davidprosser/2025/05/29/governance-start-ups-boom-in-the-battle-to-keep-ai-honest/", "date": "May 29, 2025" }, { "title": " Unbound raises $4 Mn in seed round led by Race Capital", "url": "https://entrackr.com/snippets/unbound-raises-4-mn-in-seed-round-led-by-race-capital-9317863", "date": "May 29, 2025" } ], "launches": [ { "id": 83204, "title": "🚀 Unbound Security - Secure Gen AI apps for enterprise", "tagline": "We help enterprises prevent data loss on consumer Gen AI apps", "body": "### TL;DR\n\n[Unbound Security](https://www.unboundsecurity.ai/) prevents leakage of sensitive information when employees use Gen AI applications. For example, we intervene and stop when someone pastes a code snippet that contains a secret key into ChatGPT while troubleshooting.\n\n# The problem\n\n1. **AI faces enterprise adoption roadblocks**: Despite Gen AI boosting productivity, many enterprises remain wary. Even in 2024, over 1 in 4 organizations have banned all AI apps, keeping them off-limits to employees.\n2. **AI app sprawl:** There are now well over 1,000 Gen AI apps, and many have questionable security practices. These cool “tools” continue to worry CSOs (Chief Security Officers) as employees are already using many of these to get work done.\n3. **Data privacy remains a top concern**: Over 55% of all prompts contain some form of Personally Identifiable Information (PII). This is not only a privacy nightmare, but also pushes enterprises out of compliance with multiple regulations such as PCI, SOX and HIPAA.\n\n# Our solution\n\n1. **AI app discovery:** We discover and catalog all AI apps in use by employees and help security teams understand the risk levels of each of these apps.\n\n ![uploaded image](/media/?type=post&id=83204&key=user_uploads/901694/8e15836b-4e62-4980-8805-1c87c21af0dd)\n\n2. **Granular application access policies:** Rather than block all apps by default, we allow enterprises to allow some trustworthy AI apps and steer users towards those. For example, if your org invested in Chat GPT enterprise, we help get the most out of that investment by encouraging people to use that instead of a random cool file summarizer on the internet.\n\n ![uploaded image](/media/?type=post&id=83204&key=user_uploads/901694/113c41b1-c9d2-4df4-b74d-b914db53680d)\n\n3. **Protect data leakage:** On the allowed AI apps, Unbound monitors all prompts and also blocks users from sharing sensitive information based on the configured AI usage policy.\n\n ![uploaded image](/media/?type=post&id=83204&key=user_uploads/901694/47fc0b0f-3d45-4d58-8684-64e85e0d8d5c)\n\n\n# Our ask and offer\n\n1. If you know a company that has blocked the usage of AI apps, we can help your friends use these apps with IT consent. Please drop us a note at [founders@unboundsecurity.ai](mailto:founders@unboundsecurity.ai).\n2. If you’d like to understand what people are using AI for within your org, and stay compliant, we are happy to offer our AI visibility platform for free to any YC company. Let’s [chat more over a meeting.](https://calendly.com/rajws/30-mins-meeting)\n\n# The team\n\n[Vignesh](https://www.linkedin.com/in/vigneshsubbiah16/) and [Raj](https://www.linkedin.com/in/rajaramsrinivasan/) worked together at Adobe on the same Engineering team for 5+ years, solving some of the most complex problems in the digital advertising space - The team built sub-millisecond latency systems that processed billions of advertising artifacts. Vignesh went on to be founding engineer at Tophatter and an early engineer at [Shogun (YC S18)](https://www.ycombinator.com/companies/shogun). Raj has built data security products for Palo Alto Networks and Imperva prior to Unbound where he observed the struggles of data security programs in large enterprises.\n\n![uploaded image](/media/?type=post&id=83204&key=user_uploads/901694/7b8b3e08-52e1-4d45-86ac-9963a62da345)\n\n", "url": "https://www.ycombinator.com/launches/Le0-unbound-security-secure-gen-ai-apps-for-enterprise", "votes": 121, "created_at": "2024-08-19T22:37:57.387Z" } ] }, { "id": 29825, "name": "Sepal AI", "slug": "sepal-ai", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/78c48b77fe0ad236d148f827dd2ba0b86450afd3.png", "website": "https://www.sepalai.com", "all_locations": "San Francisco, CA, USA", "long_description": "Sepal is a data research company on a mission to advance human knowledge and capabilities through safe AI.\r\n\r\nWe partner with the world’s leading AI labs and enterprises to help their models get better at the tasks people actually want them to do.\r\n\r\nWe’ve built a Cloud-Native Agent Dataset Factory which turns the process of generating evaluation and training data from manual, inconsistent, and labor-intensive into something automated, standardized, and scalable.\r\n\r\nSepal AI was founded in 2024 by engineers and operators from Vercel and Turing. We went through Y Combinator, raised several million dollars from leading investors, and already count multiple Fortune 500s and top AI research labs as paying customers.", "one_liner": "Data Development for Advanced AI", "team_size": 15, "industry": "B2B", "subindustry": "B2B", "launched_at": 1722966351, "tags": [ "AIOps", "Reinforcement Learning", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2024", "status": "Acquired", "industries": [ "B2B" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/sepal-ai", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2024/sepal-ai.json", "year_founded": 2024, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/sepalai/", "x_url": "https://x.com/sepal_ai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/78c48b77fe0ad236d148f827dd2ba0b86450afd3.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Tyler Bosmeny", "url": "https://www.ycombinator.com/people/tyler-bosmeny" }, "founders": [ { "user_id": 1996512, "full_name": "Robi Lin", "title": "Founder", "founder_bio": "Co-Founder @ Sepal AI\n\nBuilt the enterprise workflow products and fulfillment strategy at Turing.com. Scaled Turing’s LLM trainer business line from 50 to 800+ onboarded developers in 5 months for foundational LLM and enterprise customers. Previously was at Bain & Co.\n", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/robert-lin1/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4668b119bdaddeda68e7781c88272985fc03575c.jpg" }, { "user_id": 860488, "full_name": "Kat Hu", "title": "Founder", "founder_bio": "Cofounder @ Sepal AI\n\nBuilt Turing’s Foundational LLM trainer business GTM. Ran orgs of 500+ AI trainers & built corresponding operations for scale. Previously was at McKinsey.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/kathleenhu/", "x_url": "https://www.twitter.com/katqhu1", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/0060040b5fd066406e4be3460a6c0d0209a59029.jpg" } ], "jobs": [], "news": [ { "title": "Sepal AI supports the launch of Anthropic's Claude 3.7 Sonnet with Pre-launch Testing", "url": "https://x.com/sepal_ai/status/1894157588142690695", "date": "May 01, 2025" } ], "launches": [ { "id": 82824, "title": "🌱 Sepal AI - Confidently deploy your AI models", "tagline": "We build benchmarks, evals, & training datasets for enterprises.", "body": "# **Tl;dr:** Sepal provides frontier data and tooling for advancing responsible AI development.\n\n# **__________________________________________________________**\n\n[Sepal AI](https://www.sepalai.com/?ref=launch_yc) is on a mission to advance human knowledge and capabilities with the responsible development of artificial intelligence.\n\n# **🧐 Responsibly advance human knowledge with AI? What does that mean?**\n\nWe believe in a world where AI advances scientific research and empowers economic growth. \n\nTo achieve that future, AI product & model builders need:\n\n1. **Golden Datasets and Frontier Benchmarking:** To iteratively measure model performance on specific use cases.\n2. **Training Data:** To improve model capabilities using fine-tuning and RLHF.\n3. **Safety / Red-teaming:** To measure and forecast the safety of LLMs before putting them out in the wild.\n\n# **__________________________________________________________**\n\n# ⚠️ **Okay, well why does it matter?**\n\nFrontier data for AI development is vital for safe deployment & scaling. However, developing this data is difficult. \n\nMost frontier data requires domain knowledge that can be hard to source and curate (e.g., finance, medical, physics, biology, etc.). Publicly available benchmarks (e.g., MMLU, GPQA, MATH, etc.) are contaminated and too general to be useful to actual product & model builders.\n\n# **__________________________________________________________**\n\n# **🌱 How do we do this?**\n\nWe’ve built Sepal AI - the data development platform that enables you to curate useful datasets.\n\n**The Platform:** We bring data generation tooling, human experts, synthetic data augmentation, and rigorous quality control into one platform so you can manage the production of high-quality datasets.\n\n![uploaded image](/media/?type=post&id=82824&key=user_uploads/860488/81dd4889-9694-4c67-8505-389edf7c3361)\n\n**Our Expert Network:** We’ve built a network of 20k+ experts across STEM and professional services (think academic PhDs, business analysts, medical professionals, marketing and finance consultants) to support campaign design & data development. \n\nSample engagements we’ve run:\n\n* **🧬 Cell and Molecular Biology Benchmark:** An original benchmark to evaluate complex reasoning across models. Produced by a team of PhD biologists from top institutions in the US.\n* **💼 Finance Q&A + SQL Eval:** A Golden Dataset to test the ability of an AI agent to query a database and produce human-expert-level answers to complex finance questions.\n* **📏 Uplift Trials & Human Baselining:** End to end support for conducting secure in-person evaluations on model performance. \n* _…. \\[insert your custom use case next?\\]_\n\n# **__________________________________________________________**\n\n# **🙏 Asks:**\n\n1. If you are building an AI application and need to measure or improve your model, or\n2. If you are a researcher at an AI lab building or evaluating models for new capabilities / risk areas, or\n3. If you’re passionate about the development of AI, AI safety, or evals in general…\n\nLet’s chat — please check us out at [www.sepalai.com](http://www.sepalai.com) \n\n# **__________________________________________________________**\n\n# **👪 Our team:**\n\n![uploaded image](/media/?type=post&id=82824&key=user_uploads/860488/a1187807-573a-4248-9181-6378d6e2b042)\n\nMeet [Kat](https://www.linkedin.com/in/kathleenhu/) (on the left), [Robi](https://www.linkedin.com/in/robert-lin1/) (in the middle), [Fedor](https://www.linkedin.com/in/f3d0r/) (on the right)!\n\nRobi and Kat previously built the technical LLM training business for Turing. Kat on the go-to-market & operations side. Robi on the product & fulfillment side. Fedor is a long-time close friend - he was an early engineer at Vercel & Newfront where he built out foundational infrastructure.\n\nSay hi: [founders@sepalai.com](mailto:founders@sepalai.com).", "url": "https://www.ycombinator.com/launches/LXs-sepal-ai-confidently-deploy-your-ai-models", "votes": 51, "created_at": "2024-08-07T18:46:33.791Z" } ] }, { "id": 30039, "name": "Raycaster", "slug": "raycaster", "former_names": [ "Atlas" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/8e3854e63389e5cabd44e4b6563fefc3bd539147.png", "website": "https://eval.raycaster.ai", "all_locations": "New York City, NY, USA", "long_description": "Raycaster began as an agentic AI platform for knowledge work, serving dozens of publicly traded companies across life sciences, medical devices, energy, and oil and gas. It has since launched a division dedicated to training and evaluating AI systems in these industries, working directly with leading AI labs and enterprises to improve their ability to perform economically valuable work.", "one_liner": "Agentic AI for knowledge work", "team_size": 6, "industry": "Healthcare", "subindustry": "Healthcare", "launched_at": 1729820914, "tags": [ "Biotech", "Data Labeling", "Enterprise Software", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Fall 2024", "status": "Active", "industries": [ "Healthcare" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/raycaster", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2024/raycaster.json", "year_founded": 2025, "city": "New York City", "country": "US", "linkedin_url": "https://www.linkedin.com/company/raycasterai", "x_url": "https://x.com/raycasterai", "fb_url": "https://www.facebook.com/raycasterai", "cb_url": "https://www.crunchbase.com/organization/raycaster", "github_url": "https://huggingface.co/datasets/raycasterai/", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/8e3854e63389e5cabd44e4b6563fefc3bd539147.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": 299659, "full_name": "Levi Lian", "title": "Founder/CEO", "founder_bio": "Raised in a four-generation medical family, Levi chose computer science to accelerate research rather than practice it. At Stanford he worked in AI labs on early foundation model applications, then shipped enterprise-ready AI at Solvvy (acquired by Zoom), Intuit and Uber AI. Levi combines deep life-science empathy with proven AI product execution to compress drug-development timelines.", "is_active": true, "linkedin_url": "https://linkedin.com/in/levi-lian", "x_url": "https://twitter.com/levilian1", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ef9e120296d16d39ce58bfe22573967a5158a27a.jpg" } ], "jobs": [ { "id": 81306, "title": "CMC Workflow Specialist (Analytical/Tech Transfer)", "url": "https://www.ycombinator.com/companies/raycaster/jobs/C4HEspJ-cmc-workflow-specialist-analytical-tech-transfer", "location": "New York, NY, US / Remote", "type": "Contract", "role": "Operations", "role_type": null, "salary_range": "$40 - $130 / hourly", "equity_range": "", "experience": "3+ years", "visa": "US citizenship/visa not required", "skills": [] }, { "id": 81307, "title": "Clinical Trial Enablement Workflow Expert", "url": "https://www.ycombinator.com/companies/raycaster/jobs/ANItObm-clinical-trial-enablement-workflow-expert", "location": "New York, NY, US / Remote", "type": "Contract", "role": "Operations", "role_type": null, "salary_range": "$40 - $130 / hourly", "equity_range": "", "experience": "3+ years", "visa": "US citizenship/visa not required", "skills": [] } ], "news": [ { "title": "Levi Lian, CEO of Raycaster, on why vertical AI is workflows first & chat last | Sacra", "url": "https://sacra.com/research/levi-lian-raycaster-vertical-ai-workflows/", "date": "Sep 30, 2025" }, { "title": "Raycaster funding & analysis | Sacra", "url": "https://sacra.com/c/raycaster/", "date": "Sep 30, 2025" }, { "title": "Raycaster: AgentX winner", "url": "https://rdi.berkeley.edu/agentx/", "date": "Jul 21, 2025" }, { "title": "Agentic AI Summit Main Stage Speaker", "url": "https://rdi.berkeley.edu/events/agentic-ai-summit", "date": "Jul 09, 2025" }, { "title": "Founder Fellowship 2025 Participants | NYCEDC", "url": "https://edc.nyc/founder-fellowship-2025-participants", "date": "Mar 06, 2025" } ], "launches": [ { "id": 116908, "title": "Biopharma Bench 0.1 – Evaluating AI on Autonomous Drug Development", "tagline": "Frontier agents struggle in drug development; open models crush on cost", "body": "**TL;DR:** Everyone is trying to use AI for drug _discovery_ (finding molecules), but 90% of actual pharma time and money gets burned in drug _development_ - people digging through messy SharePoint, Veeva, fixing conflicting batch records (often scanned), updating clinical protocols, and answering FDA audit letters with 20 redline versions and siloed email threads with colleagues.\n\n![uploaded image](/media/?type=post&id=116908&key=user_uploads/299659/82cec064-a13e-4bfc-b527-31b46388e50c)\n\nWe built **Biopharma Bench 0.1**: 12 complete, reconstructed drug companies, 71 multi-hour tasks, and 752 ground-truth checks audited by industry veterans to see what happens when you actually put frontier AI into an employee’s seat.\n\n> One interesting finding: DeepSeek V4.1 Flash matched GPT-6 Sol on the actual work (57.0% vs 54.3%) at a sixth of the cost ($0.11 vs $0.64 per task). Meanwhile Claude Opus 5, #2 overall at 63.8%, didn't fully pass a single task.\n\nCheck out the interactive traces & leaderboard: \\\nRead the deep-dive blog post: \\\nJoin the discussion on Hacker News: \\\nX: \\\nLinkedIn: \n\n### Backstory\n\nHey everyone! We’re the team behind Raycaster ().\n\nWe first started building AI agents for regulated industries like biopharma because we see the inevitable wave of knowledge work getting handled by AI systems with human oversight (definitely not fear-mongering, but in a super exciting way!).\\\n\\\nOver the last few months, our major focus has been helping frontier AI labs globally hillclimb capabilities in underserved industries such as biopharma, energy, data centers, among others.\n\nVery quickly, we noticed something broken: almost all bio AI benchmarks test multiple-choice questions or molecular physics. Nobody was testing whether an AI agent could actually do the drug development work, open a folder of 50 disorganized PDFs and spreadsheets, find where two lab reports contradict each other, and draft an FDA-defensible document.\n\nSo, we spent the last few months working directly with former pharma directors and regulatory specialists to build an evaluation that tests what real work actually looks like.\n\n---\n\n### The Problem: The \"Fluency Trap\"\n\nIt takes 8–10 years and over $1B to get a drug to patients. Most of that delay is paperwork and coordination mistakes:\n\n* Protocol Drift: \\~75% of clinical trials have to amend their protocols mid-flight. Each amendment costs hundreds of thousands of dollars and months of lost time.\n* Messy Records: A tiny discrepancy between a stability test memo from last Tuesday and an executed batch record can trigger FDA warning letters (Form 483s) or clinical holds.\n\nWhen people play with LLMs for this, they fall into what we call the fluency trap: Frontier models write seemingly true prose. If you just read the output, the AI sounds like a 20-year VP of Regulatory Affairs who knows everything.\n\nBut when you actually grade the facts against ground truth, the model cited an outdated 2023 draft instead of the approved amendment, or hallucinated a purity threshold. In biopharma, being 95% right with one made-up number is an instant rejection.\n\n![uploaded image](/media/?type=post&id=116908&key=user_uploads/299659/46c7d62a-3842-4e51-8680-ff7f80f37255)\n\n### How Biopharma Bench Works\n\nWe didn’t give the models clean, cherry-picked prompts. We dropped them into messy environments:\n\n* 12 Companies: Real reconstructed company drives packed with thousands of messy files and software states - unredacted study reports, supplier quotes, assay validations, and random internal memos.\n* 71 Long-Horizon Tasks: Multi-hour workflows across Preclinical, CMC/Manufacturing, Clinical Ops, and Regulatory Affairs (like investigating an out-of-spec stability result or drafting a protocol amendment).\n* 752 Ground-Truth Checks: Graded against falsifiable rubrics created using real documents along with former pharma directors.\n\n![uploaded image](/media/?type=post&id=116908&key=user_uploads/299659/36d4376d-59ed-4eba-9ede-d93930299e98)\n\n---\n\n### What Shocked Us from the Runs\n\nWe ran frontier reasoning models across multiple popular harnesses (Codex, Pi, Claude Code). Here are the 3 biggest takeaways:\n\n1. **DeepSeek matched GPT-6 Sol at 1/6 the cost.** \n2. **Almost nobody finishes a full task cleanly.** Top systems get 45–70% of checks, but full passes are rare: GPT-6 Astra 8/71, Grok 4.6 2/71, Kimi K3 and GPT-5.6 Sol 1 each. Claude Opus 5, DeepSeek, GPT-6 Sol, GLM, and Gemini got 0.\n3. **The writing is fine; the sourcing isn't.** In one public task, Claude Opus 5 did everything right on the surface: it navigated the folders, read scanned protocols page by page, and wrote a clean brief. But a secondary report claimed a \"protocol-defined 30-minute window\" that doesn't exist in the controlling protocol. Opus repeated it as fact in the brief for the inspector. An experienced engineer would have flagged it.\n\n---\n\n### See the Traces\n\n![uploaded image](/media/?type=post&id=116908&key=user_uploads/299659/10059154-ce98-41cb-b0ed-f23ac8942580)\n\n---\n\n### How We Can Help\n\n* Naturally as a data / research lab, we have tons of insights into benchmark design - we recently became a proud sponsor of [evalscience.org](https://evalscience.org), a 501(c)(3) nonprofit focused on scientific agent evaluation methods and independent benchmark verification. If there’s interest, happy to share more regularly on BookFace\n* We work with / fund academics and researchers and have landed joint papers in NeurIPS, NAACL and all the big conferences. Bring your idea to us and see what can come out of it\n* Finally, if you’re a neo-lab reading this, we have lots of ready to license data, environments and evals on some of the most interesting yet underserved industries 😄 , talk to us at [team@raycaster.ai](mailto:team@raycaster.ai)", "url": "https://www.ycombinator.com/launches/UPc-biopharma-bench-0-1-evaluating-ai-on-autonomous-drug-development", "votes": 15, "created_at": "2026-09-25T17:16:44.511Z" }, { "id": 95955, "title": "Raycaster - Cursor for regulatory documents in drug development", "tagline": "AI workspace for high-stakes documents, not just code.", "body": "Hey YC 👋 I’m Levi, founder of Raycaster.\n\n### TL;DR\n\n* Raycaster is **Devin/Cursor, but for life sciences documents instead of code.**\n* It does **project-level drafting** across hundreds of PDFs, Word, Excel, and PowerPoints - not just “write a file from scratch.”\n* We’re starting with **drug development docs** (clinical, CMC, regulatory, quality) where delays and mistakes are insanely expensive.\n\n**Video (1 min):** [link](https://youtu.be/FhpC9dyBUF4)\n\n---\n\n### The problem (in plain English)\n\nIn drug development, everything important lives in documents:\n\n* protocols, reports, validation plans\n* batch records, specs, methods\n* Module 3, labels, responses to regulators\n\nThey’re scattered across SharePoint, Veeva, email, vendor portals, etc. When something changes - dose, batch size, impurity limit - you’re supposed to update _dozens_ of dependent docs so nothing contradicts anything else.\n\nToday that means:\n\n* **Research:** hunting across PDFs, Word, Excel, PowerPoints and public guidances for the right paragraph or table\n* **Draft:** copy-pasting into new docs and hand-editing boilerplate that “mostly” matches\n* **Edit:** reviewers tracking changes by eye and leaving 200 comments per document\n* **Version control:** folders called `FINAL_v7_REALLY_FINAL_clean(2).docx` and hoping nobody sends the wrong one to the agency\n\nThis happens across _every_ stage of drug development - nonclinical → clinical → CMC → labeling. When documents drift, you get delay letters, extra review cycles, or a straight-up rejection. It’s one of the big reasons timelines slip.\n\nChatGPT / Copilot can draft a single file. They don’t understand a **cross-document project** with real regulatory consequences.\n\n---\n\n### What we built\n\nRaycaster is an **AI workspace for document projects**, starting with regulatory and CMC work.\n\nThink of a repo, but for your dossiers:\n\n* A **workspace** that holds your source docs (guidances, prior submissions, lab reports, CoAs, batch records…).\n* A graph of how they depend on each other (this spec feeds this batch record feeds this Module 3 section, etc.).\n* An AI agent that can **research, draft, edit, and reconcile** across all of them—with citations and review controls.\n\nConcretely:\n\n1. **Research, with real citations**\\\n Ask: “What does EMA say about a 5× batch size increase?” or “Which CoAs support these new commercial lots?”\\\n Raycaster reads mixed formats (PDF, DOCX, XLSX, PPTX), pulls out the relevant paragraphs/tables, and shows you the answer with pinned citations back to the exact page and cell.\n\n ![uploaded image](/media/?type=post&id=95955&key=user_uploads/299659/18f9f9a3-3ea8-49c7-b50f-17df326e526e)\n\n2. **Project-level drafting**\\\n Instead of “write me a cover letter,” you say “update the batch formula + cover letter + justification for the 5× scale-up.”\\\n Raycaster drafts the actual Word docs in your workspace, using the research + your templates, and wires in all the cross-references.\n\n ![uploaded image](/media/?type=post&id=95955&key=user_uploads/299659/a80d77f1-fbde-429f-b80d-30153df19a1e)\n\n3. **Structured editing**\\\n Editing is usually harder than drafting: “bump the batch size everywhere, but don’t touch the control batch,” “update only the EMA version,” etc.\\\n Raycaster does **diff-aware edits**: it proposes precise changes inline, shows red/green diffs, and explains _why_ each change was made.\n\n ![uploaded image](/media/?type=post&id=95955&key=user_uploads/299659/4b1ab418-fbe1-4517-a157-f54b076a9ca5)\n\n4. **Version control + remote agents**\\\n Every change is tracked like a commit. You see: who asked for it, what files were touched, what the diff was, and what sources were used.\\\n You can let Raycaster run “in the background” on a workflow (e.g. keep Module 3 in sync with new CoAs), then come in as a human reviewer to approve or tweak before anything is finalized.\n\nAll of this lives in a collaborative, access-controlled workspace, so legal/reg affairs/IT don’t freak out.\n\n---\n\n### Why this is technically interesting\n\nFor devs: this is all the fun of agents and tooling, but on **documents instead of code**:\n\n* Retrieval over **huge, messy, mixed-format corpora** (PDF/DOCX/XLSX/PPTX, scanned tables)\n* A maintained **dependency graph** of citations and document links, so the agent can do impact analysis instead of one-off Q&A\n* A diff engine that understands “regulatory text” and tables (git diff , not just lines of code\n* Multi-agent workflows with strict evaluation + human checkpoints, because “just let the model commit” is not an option when the FDA is on the other side\n\n---\n\n### Why we’re doing this\n\nI started Raycaster after seeing the same pattern over and over: the science was fine; **the documents were the bottleneck.** Teams were shipping “AI pilots,” but the real work was still people copying text between PDFs and Word.\n\nWe think documents deserve the same treatment code got:\\\n**good tooling, real version control, and smart agents that understand context.**\n\n---\n\n### Ask\n\nWe’re opening prioritized access for:\n\n* **Biopharma, med-tech, and device companies** with heavy document work (regulatory, CMC, clinical, quality, safety) on major submissions or post-approval changes.\n* **Engineers / researchers** who care about serious agentic systems on messy document workflows and want to kick the tires or collaborate on evals/benchmarks.\n\nIf that’s you (or someone you know):\n\n👉 Email [**founders@raycaster.ai**](mailto:levi@raycaster.ai) with the subject **“YC Pilot”** and 2-3 sentences about your team + doc stack, or just ping me at [**raycaster.ai**](http://raycaster.ai).\n\n![uploaded image](/media/?type=post&id=95955&key=user_uploads/299659/474ffc8a-305f-4986-ba5e-b236cf06dd90)\n\n", "url": "https://www.ycombinator.com/launches/Oxf-raycaster-cursor-for-regulatory-documents-in-drug-development", "votes": 19, "created_at": "2025-12-03T18:56:35.429Z" } ] }, { "id": 30289, "name": "Sherpa Labs", "slug": "sherpa-labs", "former_names": [ "DagHub.ai" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/9fc6e69fe5c9b9ed32dbb15b81f5057c0f2e4abd.png", "website": "https://www.sherpalabs.ai", "all_locations": "San Francisco, CA, USA", "long_description": "", "one_liner": "Frontier Data", "team_size": 5, "industry": "B2B", "subindustry": "B2B", "launched_at": 1737526775, "tags": [ "Reinforcement Learning", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2025", "status": "Active", "industries": [ "B2B" ], "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/sherpa-labs", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2025/sherpa-labs.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/sherpalabs-ai", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/9fc6e69fe5c9b9ed32dbb15b81f5057c0f2e4abd.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Tom Blomfield", "url": "https://www.ycombinator.com/people/tom-blomfield" }, "founders": [ { "user_id": 2275056, "full_name": "Justin Jiang", "title": "Founder", "founder_bio": "Building frontier RL data pipelines @ Sherpa Labs. Ex-research scientist from Surge, ex-quant researcher from Millennium, Vatic, Optiver, working on medium frequency statistical arbitrage and high frequency trading algorithms. Data scientist at FedEx and Tradeweb. Graduated from Harvey Mudd with a double BS in Math and CS.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/justinjiang12/", "x_url": "https://x.com/jostinjiang", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/e4bfe9f614fc52837827a9aa23efb42eddd4c4af.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 30303, "name": "AfterQuery", "slug": "afterquery", "former_names": [ "Cronus" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/b63e52a3ec831660a5917dbb85f52cfd61f714e9.png", "website": "https://afterquery.com", "all_locations": "San Francisco, CA, USA", "long_description": "AfterQuery is an applied research lab curating data solutions for frontier foundation model development. Serving every frontier AI lab.", "one_liner": "Applied research lab curating data solutions for foundation model…", "team_size": 30, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1738219847, "tags": [ "B2B", "Data Labeling", "Big Data", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Winter 2025", "status": "Active", "industries": [ "B2B", "Infrastructure" ], "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/afterquery", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2025/afterquery.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/afterquery", "x_url": "https://x.com/AfterQuery", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/b63e52a3ec831660a5917dbb85f52cfd61f714e9.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Gustaf Alstromer", "url": "https://www.ycombinator.com/people/gustaf-alstromer" }, "founders": [ { "user_id": 2382342, "full_name": "Spencer Mateega", "title": "Founder/CEO", "founder_bio": "Building AfterQuery. Previously interned in technology private equity at Silver Lake, technology investment banking at Morgan Stanley, and software engineering at Meta and Google.\n\nB.S. in Finance and Statistics at Wharton and a M.S. Computer Science at Penn.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/spencermateega", "x_url": "https://x.com/spencermateega", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/afbc3ee1379a69d69496a3aa75ca799fcd63a8fd.jpg" }, { "user_id": 503252, "full_name": "Carlos Georgescu", "title": "Founder/CTO", "founder_bio": "Building AfterQuery. Previously interned at Citadel Securities (2x), Meta & Google as a software engineer. Past acquired ed-tech founder. \n\nStudied for a B.S. in Computer Science from the University of British Columbia. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/carlossgg/", "x_url": "https://x.com/CarlosGeorgescu", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/7e75fffea45bad544c5ba2ee3ea005b9731dda68.jpg" } ], "jobs": [ { "id": 75196, "title": "Software Engineer Intern", "url": "https://www.ycombinator.com/companies/afterquery/jobs/6a7szXP-software-engineer-intern", "location": "San Francisco, CA, US / Remote (US)", "type": "Internship", "role": "Engineering", "role_type": "Frontend", "salary_range": "$7.5K - $10K / monthly", "equity_range": "", "experience": null, "visa": "US citizen/visa only", "skills": [] }, { "id": 82771, "title": "Software Engineer - Applied AI / Platform", "url": "https://www.ycombinator.com/companies/afterquery/jobs/7tsT0ks-software-engineer-applied-ai-platform", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$140K - $200K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [ "Node.js", "Python", "Next.js" ] }, { "id": 87019, "title": "Software Engineer - Hiring Infra (LatAM)", "url": "https://www.ycombinator.com/companies/afterquery/jobs/PohDLWh-software-engineer-hiring-infra-latam", "location": "San Francisco, CA, US / Remote (BR; CL; MX; AR; CO; PE)", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$7K - $10K / monthly", "equity_range": "", "experience": "3+ years", "visa": "US citizenship/visa not required", "skills": [] }, { "id": 91403, "title": "Strategic Projects Lead", "url": "https://www.ycombinator.com/companies/afterquery/jobs/iTt13cJ-strategic-projects-lead", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$200K - $400K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 91406, "title": "Technical Strategic Projects Intern", "url": "https://www.ycombinator.com/companies/afterquery/jobs/8NagVRQ-technical-strategic-projects-intern", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$90K", "equity_range": "", "experience": "", "visa": "US citizen/visa only", "skills": [] }, { "id": 91408, "title": "Strategic Projects Associate", "url": "https://www.ycombinator.com/companies/afterquery/jobs/qwNKclV-strategic-projects-associate", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$200K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 91409, "title": "Strategic Projects Intern", "url": "https://www.ycombinator.com/companies/afterquery/jobs/Ep6eNax-strategic-projects-intern", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$90K - $90K", "equity_range": "", "experience": "", "visa": "US citizen/visa only", "skills": [] }, { "id": 91415, "title": "Technical Strategic Projects Associate", "url": "https://www.ycombinator.com/companies/afterquery/jobs/ziJE7ZY-technical-strategic-projects-associate", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$200K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 91417, "title": "Marketing Lead", "url": "https://www.ycombinator.com/companies/afterquery/jobs/Ww01atn-marketing-lead", "location": "San Francisco", "type": "Full-time", "role": "Marketing", "role_type": null, "salary_range": "$160K - $220K", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 91418, "title": "Technical Strategic Projects Lead", "url": "https://www.ycombinator.com/companies/afterquery/jobs/t0uj8SF-technical-strategic-projects-lead", "location": "San Francisco", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$200K - $400K", "equity_range": "", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [] }, { "id": 91613, "title": "Senior Software Engineer, Infrastructure & Platform", "url": "https://www.ycombinator.com/companies/afterquery/jobs/TU2SAvz-senior-software-engineer-infrastructure-platform", "location": "San Francisco, CA, US", "type": "Full-time", "role": "Engineering", "role_type": "Full stack", "salary_range": "$220K - $280K", "equity_range": "", "experience": "6+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "AfterQuery Becomes YC’s Fastest Unicorn At $3.2 Billion", "url": "https://www.forbes.com/sites/annatong/2026/09/01/afterquery-becomes-ycs-fastest-unicorn-at-32-billion/", "date": "Sep 01, 2026" }, { "title": "This 23 Year-Old’s New AI Data Company Has Already Hit A $100 Million Run Rate", "url": "https://www.forbes.com/sites/annatong/2026/04/09/this-23-year-olds-new-ai-data-company-has-already-hit-a-100-million-run-rate/", "date": "Apr 09, 2026" } ], "launches": [ { "id": 87549, "title": "AfterQuery - High-quality AI starts with high-quality human data", "tagline": "Powering AI models with expert-level human data", "body": "## **TL;DR**\n\nAI models need high-quality training data to effectively automate complex professional tasks - we provide that data. If you are building an AI model or need fine-tuning data for your agent, we’d love to help. \n\n\\------------\n\n## **Why We’re Building AfterQuery**\n\nThere is a dearth of high-quality training data.\n\nCurrent AI models are trained to be smart but are not trained on the work of actual professionals in many fields. \n\nFor example, current models are largely unusable to professionals working in finance (private equity, hedge funds, investment banks, etc). See our [paper](https://arxiv.org/pdf/2501.18062), where LLMs fail 60% of realistic daily tasks – but with some fine-tuning on high-quality training data, significantly improve.\n\n![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeJndLBLBOQJVpf7yPJNA2WFddpffkxqdE1rFqmEfxo3FRmcSCVWqHnsOR793XuN1iCUm-ibeWfFj-FV_3FaswWhXetsEhembFLjln5Lwc9LA0tJV6uNq7MPWyjdDq7FTZSRdmekQ?key=_udDbu5NRw5fwLzMYyqM9i52)\n\n_Chart showing LLM performance against FinanceQA_\n\nAnd the lack of high-quality training data is everywhere:\n\n* Domain-specific AI (legal, consulting, healthcare, etc.)\n* Multimodal data (audio, multilingual, etc.)\n* Foundational models looking to push the boundary on reasoning and grading\n\n## **Solution**\n\nWe are building a platform to onboard experts to work on custom datasets. Think working professionals at top firms and the smartest students in various subjects (PhDs). Then, whenever you have a request for a certain type of data, we will construct and lead a team to produce:\n\n* Post-training / fine-tuning data\n* Human-Led Reinforcement Learning (RLHF) data\n* Custom benchmarks\n\nYou will know the people making your dataset and their qualifications while we coordinate with you to ensure quality standards. We save you time and energy, so you can focus on the model and not the data.\n\n## **Example Datasets We Provide**\n\n_Foundational Model Advancing Coding Capabilities:_\n\n* New data structures and algorithms problems with test cases and step-by-step answer reasoning \n* Computer architecture and system design reasoning\n* Performance optimization code with benchmarks and reasoning\n\n_Enterprise SaaS Company Building Internal AI Developer Tool:_\n\n* Codebase-wide code snippets with corresponding refactoring suggestions and explanations \n* System and runtime errors paired with step-by-step debugging solutions \n* Database query optimization solutions and reasoning steps\n\n_Startup Building an AI Agent for Law:_\n\n* Answered legal questions with expert reasoning\n* Computer use data for common tasks on legal project management software\n* Contract redlines with expert corrections\n\n**If AI is one day to replace jobs entirely, it can’t just be good, it needs to be near perfect. This transformation will require an immense amount of high-quality training data.**\n\n## **Our Ask**\n\n* Are you **working on a foundational model** advancing multimodal applications, nuanced reasoning, or domain-specific expertise?\n* Are you **implementing AI in a specialized domain** where current models lack expertise (e.g., medicine, legal, finance, software, government, healthcare)?\n\nWe would love to meet you! \n\n## **Our Team**\n\nCarlos and Spencer first met in high school at a summer program at Google and then interned together at Meta. Danny and Spencer met in high school, where they built a startup. They then both got into Wharton, sold that startup, and became roommates.\n\n[Carlos](https://www.linkedin.com/in/carlossgg/) has built multiple software businesses on his own, exited an ed-tech startup, and has worked as a software engineer at Citadel Securities. He’s won the largest possible academic scholarship in all of Canada, competed in state and national sailing competitions, and has 8x his bankroll in poker.\n\n[Spencer](https://www.linkedin.com/in/spencermateega) wrote an award-winning research paper and interned at a private equity firm in high school, before deciding to intern at Meta (SWE), Google (SWE), Morgan Stanley (IB), and Silver Lake (PE) in the next 3 years. He was the sole summer analyst at Silver Lake globally and has programmed numerous apps in his spare time (and sold some of them). He trains for half marathons and lifts in his free time between completing a master’s in computer science concurrently with his Wharton undergraduate degree.\n\n[Danny](https://www.linkedin.com/in/danny1898/) competed and won international competitions in public speaking during high school while pursuing his interest in AI, winning an international Microsoft AI competition and conducting research at a machine learning neuroscience lab. He then learned how people build real estate, from grocery stores to billion-dollar luxury hotels, eventually working on the largest real estate IPO in history and the largest IPO in 2024. He also really likes movies and attended the world's largest film festival, catching premieres of films like Anora and The Substance.\n\n![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXfkN7pvcMyTHjf5fZe7qJrF9xiyAvI0yAJ0PHNDdxWcOcU0gw5cDFUzOq20x68YJeb40Pk8dPXsNGsPVSnhYNZu3AHEAoZHcSTQ3nYyLYvMNcthHoAkoERsmzARwkGFS-fRnslLNw?key=_udDbu5NRw5fwLzMYyqM9i52)\n\n", "url": "https://www.ycombinator.com/launches/Mm5-afterquery-high-quality-ai-starts-with-high-quality-human-data", "votes": 114, "created_at": "2025-02-10T21:58:10.460Z" } ] }, { "id": 30500, "name": "Besimple AI", "slug": "besimple-ai", "former_names": [ "Simple Annotation", "besimple ai" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/8305d688995d7c8f184d463f02f3ded6947c12e7.png", "website": "https://besimple.ai", "all_locations": "San Francisco, CA, USA", "long_description": "We are building the data layer for AI, starting with audio. \r\n\r\nWe start with data collection, curating our own proprietary set of diverse conversational data covering a wide range of languages, dialects and accents. We then leverage human expert audio annotators and our own annotation platform to process audio data for Automatic Speech Recognition. \r\n\r\nWith human level transcription and diarization, our data help push the audio model frontier. Today we have over millions of hours of conversational data, and growing. \r\n\r\nIf you need audio data for training or evaluating your voice models or voice agents, reach out! We offer flexible licensing deals that work for startups and enterprises, with minimal process. \r\n\r\nAudio data should besimple :) ", "one_liner": "Voice data for AI", "team_size": 10, "industry": "B2B", "subindustry": "B2B", "launched_at": 1747861843, "tags": [ "AIOps", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Spring 2025", "status": "Active", "industries": [ "B2B" ], "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/besimple-ai", "api": "https://devasheeshg.github.io/yc-api/batches/spring-2025/besimple-ai.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/besimple-ai/", "x_url": "https://x.com/besimple_ai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/8305d688995d7c8f184d463f02f3ded6947c12e7.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/14e2e0920e736148a86789dc8875ba75c3eb4e88.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/0009a9ebc2ee89017db7e4d65efbcee0f22e1954.png" ], "primary_partner": { "name": "Nicolas Dessaigne", "url": "https://www.ycombinator.com/people/nicolas-dessaigne" }, "founders": [ { "user_id": 2527745, "full_name": "Yi Zhong", "title": "Founder/CEO", "founder_bio": "Currently co-founder & CEO at Besimple AI, building the human-interaction data layer for emotionally intelligent AI. \n\nPrior to Besimple, I was a PM lead at Meta for its AI Integrity team, building large scale AI systems to detect problematic contents and accounts. I also worked as a PM at Microsoft and Dropbox, after graduating from MIT in 2016.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/yzhong94/", "x_url": "https://x.com/yiz_be_building", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/efec15c007a345ab4e139c02c9ec73a69f2fb207.jpg" }, { "user_id": 2544844, "full_name": "Bill Wang", "title": "Founder/CTO", "founder_bio": "Shenzhen-born, Rhode Island-brewed, Bay Area-domesticated. While at Meta, launched 7 products and killed 2. Most recently, led the GenAI Annotation team, developing an in-house annotation platform for training LLaMa. Previously, managed an engineering organization responsible for improving connectivity for over 300 million users and optimizing 70%+ of Meta's annual SMS spend.\n", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/8illwang/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/76fe485d198fa8330635f73a2b97cb0314d66721.jpg" } ], "jobs": [ { "id": 96880, "title": "Audio QA Lead - Part Time Contractor", "url": "https://www.ycombinator.com/companies/besimple-ai/jobs/mKH99fq-audio-qa-lead-part-time-contractor", "location": "San Mateo, CA, US / Remote (US)", "type": "Contract", "role": "Operations", "role_type": null, "salary_range": "$25 - $45 / hourly", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 97432, "title": "Strategic Projects Lead — Audio Data", "url": "https://www.ycombinator.com/companies/besimple-ai/jobs/yWfhhOR-strategic-projects-lead-audio-data", "location": "San Mateo, CA, US", "type": "Full-time", "role": "Operations", "role_type": null, "salary_range": "$140K - $160K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 110630, "title": "Community Growth Manager, Contributor Acquisition", "url": "https://www.ycombinator.com/companies/besimple-ai/jobs/pDSSREY-community-growth-manager-contributor-acquisition", "location": "San Mateo, CA, US / Remote (US)", "type": "Contract", "role": "Operations", "role_type": null, "salary_range": "$30 - $50 / hourly", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 116834, "title": "RL Environment Engineer (Full-time, Part-time, or Contract)", "url": "https://www.ycombinator.com/companies/besimple-ai/jobs/dPIlkr6-rl-environment-engineer-full-time-part-time-or-contract", "location": "San Mateo, CA, US / Remote (US)", "type": "Full-time", "role": "Engineering", "role_type": "AI engineer", "salary_range": "$160K - $360K", "equity_range": "", "experience": "Any (new grads ok)", "visa": "US citizen/visa only", "skills": [] }, { "id": 116900, "title": "Linguist - Speech & Conversational AI", "url": "https://www.ycombinator.com/companies/besimple-ai/jobs/uTWqXak-linguist-speech-conversational-ai", "location": "San Mateo, CA, US / Remote (US)", "type": "Contract", "role": "Operations", "role_type": null, "salary_range": "$30 - $60 / hourly", "equity_range": "", "experience": "3+ years", "visa": "US citizen/visa only", "skills": [] } ], "news": [ { "title": "Announcing our seed round led by YCombinator - Besimple AI", "url": "https://besimple.ai/blog/announcing-our-seed", "date": "Dec 05, 2025" } ], "launches": [ { "id": 90653, "title": "Besimple AI – Spin up your own data annotation platform in 60 seconds🚀", "tagline": "We make data annotation simple", "body": "Hey there! We are [Yi](https://www.linkedin.com/in/yzhong94/) and [Bill](https://www.linkedin.com/in/8illwang/), from [Besimple AI](https://www.ycombinator.com/companies/besimple-ai)! Struggling with data annotation? Keep reading!\n\n💎 **TLDR**\n\nWith Besimple, you can generate your own platform for annotating AI eval and training data instantly.  Simply paste or stream any type of raw data and we'll instantly generate a tailored annotation interface, clear guidelines, an automated human-in-the-loop workflow, and AI judges to scale your insights.\n\nhttps://www.youtube.com/watch?v=Hw-EVKdOhds\n\n## **❌ The Problem**\n\nHigh-quality, human-reviewed data is essential for improving AI models. But teams today face significant challenges:\n\n* **Complex LLM Workflows**: Most AI startups and even the largest AI companies use spreadsheets for annotations to “move fast”. It is brittle. And it breaks down quickly for dynamic, multimodal, agent-based LLM use cases.\n* **Annotation Bottlenecks**: As models get better, only domain experts and someone on your team can improve model performance.  They quickly get overwhelmed when guidelines and data change constantly. This is expensive and slows model iteration.\n* **No Feedback Loop**: After a model ships, interesting production data isn’t automatically fed back into evaluation or training, missing opportunities for ongoing improvement.\n\n## **✅ Our Solution:** \n\n1. Instant custom UI: Just paste or stream your data and we create a custom annotation UI for your task.  We support text, chat, audio, video, LLM traces, and more. \n2. Tailored guidelines: Import any existing guidelines or we will draft new ones aligned with your business goals, ready for annotation\n3. AI Judges for real-time evaluation: LLM-based “judges” continuously learn from incoming annotations to evaluate live traffic and flag borderline cases for human review.\n4. Enterprise-grade deployment: On-prem optional install, and robust user management for internal SMEs, external vendors, or Besimple’s vetted annotators.\n5. Lightning fast setup: No code, no plugins—just drop in your data, set guidelines, and you’re good to go.\n\n**Traction:** The leading AI grading company Edexia uses besimple to annotate hundreds of decisions and improve their evals.\n\n## **The Team**\n\nWe’re **Yi Zhong** and **Bill Wang** —who built the annotation platform for Meta’s Llama models. We founded Besimple AI to help everyone spin up a Scale AI in 60s.\n\n# **🙏 Our asks**\n\nWe’re on a mission to make _continuous, high-quality data_ the easiest part of AI development. If that resonates, let’s chat!\n\n1. Reach us at [**founders@besimple.ai**](mailto:founders@besimple.ai)\n2. Book demo [here](https://calendly.com/yi-besimple/besimple-ai-intro-call).\n3. Or join our self-serve waitlist on our [website](https://besimple.ai/)!\n\nWe’ll show you how to launch a custom annotation platform in under a minute!", "url": "https://www.ycombinator.com/launches/Na9-besimple-ai-spin-up-your-own-data-annotation-platform-in-60-seconds", "votes": 332, "created_at": "2025-05-22T20:14:47.012Z" } ] }, { "id": 30501, "name": "Cartpole", "slug": "cartpole", "former_names": [ "Prophet", "Jazzberry" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/154693ac9eca92a2d8f93ead7153a0858c38d2b3.png", "website": "https://cartpole.com", "all_locations": "San Francisco, CA, USA", "long_description": "We're creating reinforcement learning environments for training frontier models.", "one_liner": "Building reinforcement learning environments", "team_size": 2, "industry": "B2B", "subindustry": "B2B", "launched_at": 1744600545, "tags": [ "Reinforcement Learning", "Data Labeling", "AI", "ML" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Spring 2025", "status": "Active", "industries": [ "B2B" ], "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/cartpole", "api": "https://devasheeshg.github.io/yc-api/batches/spring-2025/cartpole.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/cartpole", "x_url": "https://x.com/cartpoleai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/154693ac9eca92a2d8f93ead7153a0858c38d2b3.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": 2557020, "full_name": "Mateo Perez", "title": "Founder/CEO", "founder_bio": "Building Cartpole. PhD @ CU Boulder.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/mateo-perez/", "x_url": "https://x.com/m_perez_", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/bc4e196cc3b6bc16eb7775444008cd50ad1ddd64.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 90047, "title": "Jazzberry - AI bug finding with real code execution", "tagline": "We're creating the best AI agent for bug finding", "body": "Hello everyone, \n\nWe’re Mateo and Marco, and we're excited to introduce Jazzberry, an AI agent for finding bugs. Jazzberry is powered by running your code in a secure sandbox and integrates directly into your pull request workflow on GitHub.\n\n**Try it here →** \n\n\n\n**Why we built Jazzberry**\n\nAI code generation has dramatically increased the amount of code that is written without the tools for bug finding keeping up, leading to a world with more buggy code that is hindering developers. To improve bug finding, we’ve created an AI agent with the ability to run snippets of your code. We're not just flagging potential issues based on code patterns; we're observing the real behavior of your software. Here’s how it works.\n\n1. When you create a pull request, Jazzberry gets to work.\n2. Your repository is cloned into a microVM.\n3. Our agent examines the contents of the pull request to guide its search. It can write and execute code in the microVM as it looks for bugs.\n4. Jazzberry returns a clear markdown table directly as a comment.\n\n![uploaded image](/media/?type=post&id=90047&key=user_uploads/2557020/31d75b00-abc3-4a60-8ef5-b3a685edfcd5)\n\n**Real Bugs, Real Impact**\n\nWe've already seen Jazzberry make a tangible difference for our early users. So far, Jazzberry has…\n\n* successfully identified a SQL injection vulnerability\n* discovered a critical logical error in API key validation\n* [found an incorrect page navigation](https://blog.openint.dev/how-we-scale-prs/) (see below, we are formerly known as Prophet)\n\n![uploaded image](/media/?type=post&id=90047&key=user_uploads/2557020/864328cf-c4b9-4b9a-b267-e9059a158f32)\n\n**Who we are**\n\nWe have been friends for over a decade, and are coming together to create Jazzberry. Marco completed his MS at the University of Stellenbosch specializing in LLMs for automated test suite generation. Mateo completed his PhD at the University of Colorado Boulder with a research focus on reinforcement learning and formal methods.\n\n**Our Ask**\n\n* **Try it out** – Jazzberry is available now with a free trial at .\n* **Contact us** – We’d love to hear your feedback and what bugs Jazzberry helps you catch. Reach us at [founders@jazzberry.ai](mailto:founders@jazzberry.ai).", "url": "https://www.ycombinator.com/launches/NQN-jazzberry-ai-bug-finding-with-real-code-execution", "votes": 42, "created_at": "2025-05-05T22:37:45.137Z" } ] }, { "id": 30561, "name": "Sureform", "slug": "sureform", "former_names": [ "Valora", "Daxa" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/aa0aa8a269ab91daa7f1007cbe3b7c767e28dc44.png", "website": "https://www.sureformhq.com/", "all_locations": "San Francisco, CA, USA", "long_description": "Sureform help companies monetize their internal operational data by turning it into training signal for frontier AI.", "one_liner": "Real-world company data for frontier AI", "team_size": 4, "industry": "B2B", "subindustry": "B2B", "launched_at": 1744274885, "tags": [ "Marketplace", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Spring 2025", "status": "Active", "industries": [ "B2B" ], "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/sureform", "api": "https://devasheeshg.github.io/yc-api/batches/spring-2025/sureform.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/sensedata-ai", "x_url": "https://x.com/sureform_us", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/aa0aa8a269ab91daa7f1007cbe3b7c767e28dc44.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": 2579059, "full_name": "Ananth Kashyap", "title": "Founder/CEO", "founder_bio": "Founder, CEO @ Sureform | Previously: Penn, Google", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/ananth-kashyap/", "x_url": "https://x.com/ananthkashyap99", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b2a387d1b7efa302de62570b9f4ae79e0eee7d6a.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 98060, "title": "Sureform: Turning physical expertise into training data for world models and robotics", "tagline": "Building the platform to acquire task-specific robotics data at scale", "body": "Sureform connects workplaces like hotels, warehouses, and restaurants with robotics labs to collect task-specific training data for world models and robot foundation models.\n\n**Problem**\n\nTo create a future where robots can reliably perform specialized tasks, like restocking inventory in a pharmacy or repairing downed power lines, robotics teams need massive amounts of task-specific human and robot data. Data that’s crafted in simulation or in controlled environments doesn’t reflect the natural variation and complexity of the world, and as a result, we need significantly more real-world action data across each of the millions of highly-specialized physical tasks that need to be automated.\n\n**Solution**\n\nTo solve this problem, Sureform’s platform enable robotics teams to get task-specific data from workplaces like hotels, construction sites, laundromats, and warehouses, where employees are equipped with head-mounted RGB cameras, stereo depth rigs, and handheld grippers that capture time-synced video, audio, pose, depth, and tactile data as they work. Every session of data captured through our platform is structured, synchronized, and delivered training-ready.\n\n\n\n\\\n**Our Ask**\n\nIf you’re working on robot foundation models or world models (or know anyone who is), reach out to [ananth@sureformhq.com](mailto:ananth@sureformhq.com) or schedule a meeting here:[ https://calendly.com/sureform-us/30min](https://calendly.com/sureform-us/30min)\n\nAnd if you own or work at a business that’s interested in growing revenue through data collection and, later down the line, automating physical ops with robots, reach out as well!", "url": "https://www.ycombinator.com/launches/PVc-sureform-turning-physical-expertise-into-training-data-for-world-models-and-robotics", "votes": 9, "created_at": "2026-02-23T20:01:00.542Z" } ] }, { "id": 30622, "name": "Liva AI", "slug": "liva-ai", "former_names": [ "Symbia", "Liva" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/6d577992606b8392c9619b8dffacec074b48668d.png", "website": "https://www.theliva.ai", "all_locations": "San Francisco, CA, USA", "long_description": "Helping to build more socially intelligent AI, starting with voice.", "one_liner": "Building socially intelligent AI", "team_size": 6, "industry": "B2B", "subindustry": "B2B", "launched_at": 1753401226, "tags": [ "Marketplace", "B2B", "Data Labeling", "Big Data", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2025", "status": "Active", "industries": [ "B2B" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/liva-ai", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2025/liva-ai.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/thelivaai/", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/6d577992606b8392c9619b8dffacec074b48668d.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Jon Xu", "url": "https://www.ycombinator.com/people/jon-xu" }, "founders": [ { "user_id": 2218085, "full_name": "Ashley Mo", "title": "Founder/CEO", "founder_bio": "Co-founder & CEO | prev. Caltech CS | Published research on an AI model that could detect lung disease through cough recordings at 17 y/o (IEEE Xplore) and led large-scale research study to collect health audio data @MIT | Built biosensors for chronic kidney disease @Caltech", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/ashley-mo/", "x_url": "https://x.com/ashlleymo", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/560a22eeec798d2a8bbad3f734a78566c4e25c11.jpg" }, { "user_id": 1393905, "full_name": "Aoi Otani", "title": "Founder/CTO", "founder_bio": "Co-founder & CTO ", "is_active": true, "linkedin_url": "https://linkedin.com/in/aoi-otani-ab3a1b205", "x_url": "https://x.com/bigvalleyblue", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4d5f89e70756e3eca913dfa01582a7032fe54671.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 92886, "title": "Liva AI – Real Voice & Video Data for AI", "tagline": "We provide real voice and video data for developing realistic AI.", "body": "**TL;DR:** We provide high-quality human voice and video datasets for companies building realistic AI models. Real data sourced in-house across diverse languages, emotions, and contexts.\n\n**Ask:** [Talk to us](https://calendly.com/ashleymo-theliva/30min) if you're a researcher or founder building voice/video generation models. We can help you.\n\nhttps://youtu.be/PwjTSZRC8XE\n\n**The Problem:**\\\nCompanies building voice and video AI models need quality training data to reach human-like performance for models. Scraping has limits and existing datasets lack diversity across accents, emotions, and languages. As AI moves toward human-like interactions like customer conversations, therapy sessions, classroom teaching, or entertainment, models need authentic human expressions that the internet can't supply.\n\n**Our Solution:**\\\nLiva collects real, consented human voice and video entirely in-house—no synthetic data or third-party purchases. All content is authentic and rights-cleared.  We’re already delivering our voice dataset to a lab training expressive foundation models for voice.\n\nWe capture diverse accents, emotional range, and varied contexts (sales calls, multi-channel dialogues, expressive monologues, casual conversations, job interviews, and more) with high production quality through crowdsourcing and strategic partnerships.\n\n**The Team:**\n\nWe’ve worked on many research projects together since we met 3 years ago.\n\n* **Ashley:** Caltech CS dropout. Built an AI model to detect lung disease from cough recordings (@ MIT) and led large-scale data collection initiatives. \n* **Aoi:** Prev. Harvard Bio/CS. Did ML research in representation learning and image diffusion models. 5+ Publications in ICML, Nature, etc.\n\n**Our Ask:**\n\nIntroductions to:\n\n* AI labs building or fine-tuning voice, video, and multimodal generation models\n* Companies using or integrating voice/video models\n* Film, audio, and content production companies\n* Researchers working on audio, video, and multimodal generative models\n\nContact us: [founders@theliva.ai](mailto:founders@theliva.ai)", "url": "https://www.ycombinator.com/launches/OAA-liva-ai-real-voice-video-data-for-ai", "votes": 50, "created_at": "2025-08-12T03:48:13.634Z" } ] }, { "id": 30899, "name": "Sciloop", "slug": "sciloop", "former_names": [ "SciLoop" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/f01d1e4f95660e72f930c4f8b2542e2224dc9107.png", "website": "https://sciloop.dev", "all_locations": "San Francisco, CA, USA", "long_description": "Sciloop creates expert-level math and physics problems that frontier AI models can't solve, then sells the data to AI labs for training and evaluation. Our problems are created by IPhO and IMO medalists — the top 0.01% of STEM talent globally. On our benchmark, models like GPT 5.4 Pro and Gemini 3.1 Pro score 0-5% on our hardest problems. We work with AI labs to supply continuous, fresh training data that pushes the frontier of mathematical and scientific reasoning. Founded by Bilal and Osman, International Physics Olympiad medalists from MIT with hands-on ML research experience at MIT CSAIL.\r\n", "one_liner": "Expert STEM reasoning data for frontier AI labs.", "team_size": 6, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1761248663, "tags": [ "Marketplace", "Data Labeling", "Big Data", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Fall 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/sciloop", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2025/sciloop.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/sciloop-labs", "x_url": "https://x.com/SciloopAI", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/f01d1e4f95660e72f930c4f8b2542e2224dc9107.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/fed7f882927e8fb06da3fa37372d4882ade38f87.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/7236dec827b854b72e6c08d373f971278c84b65f.png" ], "primary_partner": { "name": "Aaron Epstein", "url": "https://www.ycombinator.com/people/aaron-epstein" }, "founders": [ { "user_id": 1933114, "full_name": "Bilal Asmatullah", "title": "Founder/CEO", "founder_bio": "Co-founder of Sciloop (YC F25), On Leave From MIT\nInternational Physics Olympiad ’22 medalist.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/bilal-asmatullah-8705692b4/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/bf61bc3516b5794d469bbefebfcaa598357a1e97.jpg" }, { "user_id": 2718484, "full_name": "Osman Siddique", "title": "Founder", "founder_bio": "CTO and co-founder of Sciloop (YC F25). \nCS student at MIT (on leave). Medalist and two-time highest national scorer at the 2022 International Physics Olympiad.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/osman-siddiquer", "x_url": "https://x.com/OsmanSiddiquer", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/65092c9b010287aa5db264ecfeb813617bfb76af.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 94951, "title": "Sciloop: The AI Scientific Researcher", "tagline": "Come up with novel hypotheses, run experiments, analyze results, and write the paper — all automatically.", "body": "**TL;DR: We’re making the AI that does your research for you.[ Book a call.](https://calendly.com/osman-sciloop/30min)**\n\nHi.\n\nWe’re building **Sciloop**, a platform that automates the entire scientific research loop from hypothesis generation to experiment design, execution, analysis, iteration, and drafting real research papers.\n\nhttps://youtu.be/GNM8iA6p9Ao\n\n**Our Story:**\n\nWe’re Bilal and Osman, IPhO medalists and former MIT students with 4 years of experience applying ML in various research domains. While working at MIT CSAIL, we saw how large language models could help with literature review and even generate new hypotheses. Stanford’s recent study ([Can LLMs Generate Novel Research Ideas?](https://arxiv.org/abs/2409.04109))  showed that LLM-generated hypotheses can be more novel than those from domain experts. We saw the same in our own work.\n\n![uploaded image](/media/?type=post&id=94951&key=user_uploads/2718484/3a420779-570b-4a69-b3ee-628644695a3e)\n\nBut after generating ideas, everything slowed down. We spent hours designing experiments, writing code, debugging compute setups, and rerunning failed jobs that could crash after 12 hours. Even with tools like Cursor that helped us code faster, the overall loop from idea to insight was still painfully manual. Researchers everywhere face this same bottleneck.\n\nSo we asked ourselves: what if the research loop itself could be automated?\n\nSo we built **Sciloop**. It takes a research intent — a question or hypothesis — and automatically designs, runs, and monitors experiments in parallel on our own managed cloud compute. It tracks metrics, analyzes results, recommends next steps, and finally drafts a paper summarizing methods and findings. Sciloop turns days of busywork into hours of progress.\n\nOur goal is simple: **to completely automate scientific discovery**. The next decade will see a century of progress packed into a few years. As AI grows smarter and more accessible, its biggest impact will be speeding up science — finding new cures, discovering better materials, and driving breakthroughs in physics and math.\n\nThat’s our north star. But we’re not waiting for the future. We’re helping ML researchers move faster **right now**, automating the parts of their workflow that slow them down most. Sciloop is already being used in MIT research groups and by early partners working on real ML problems.\n\nIf you’re a company, research lab, or ML team struggling with model performance, experimentation speed, or scaling research, we’d love to work with you. Our goal is to deliver concrete improvements on your research challenges today.\n\n🚀 **The Ask - Let’s Make Research Self-Improving!**\n\nIf you’re working on ML research that could move faster, **book a quick chat** at https://calendly.com/osman-sciloop/30min or email us at [**team@sciloop.dev**](mailto:team@sciloop.dev).\n\nWe help brilliant research ideas stop gathering dust. This is for anyone who:\n\n* Runs **university or industry ML labs** drowning in experiments.\n* Is a **research student** buried under mountains of code and papers.\n* Leads a **company dreaming of R&D** but lacking scientists or resources.\n* Has a **head full of creative ideas** but no time to bring them to life.\n\nLet’s collaborate and accelerate your research—together.\n\n\\[1\\][ Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers](https://arxiv.org/abs/2409.04109)", "url": "https://www.ycombinator.com/launches/OhT-sciloop-the-ai-scientific-researcher", "votes": 33, "created_at": "2025-10-29T01:25:09.249Z" } ] }, { "id": 31223, "name": "Datoric", "slug": "datoric", "former_names": [ "Flow AI", "Arzule" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/fa894803120df3a3a9250c66e4232e33e2a223d0.png", "website": "https://www.datoric.com", "all_locations": "San Francisco, CA, USA", "long_description": "Datoric develops custom datasets for voice models, robotics, and world models, treating research, collection, verification, and production as one continuous process. We work closely with frontier model teams to turn emerging limitations and model failures into testable data hypotheses, while running our own experiments ahead of customer demand. This allows us to operationalize validated methods into repeatable collection systems at scale.\r\n\r\nData is collected through private invite-only applications separated by modality, customer, and trust level. Every submission remains linked to the contributor, device, task, session, consent, rights, and processing history that produced it, giving our internal QA and fraud models the context to detect problems that may appear legitimate in the finished file. Each collection reveals new failure cases and quality signals that improve the systems behind the next dataset.\r\n\r\nOnce a collection method is validated, we scale it up as a solution to the model failure and it also becomes a reusable data recipe for future custom projects or independently collected, rights-cleared data products. ", "one_liner": "The secure training data R&D engine for physical AI", "team_size": 2, "industry": "B2B", "subindustry": "B2B", "launched_at": 1767154035, "tags": [ "Machine Learning", "Robotics", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "status": "Active", "industries": [ "B2B" ], "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/datoric", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/datoric.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/arzulehq/", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/fa894803120df3a3a9250c66e4232e33e2a223d0.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Harshita Arora", "url": "https://www.ycombinator.com/people/harshita-arora" }, "founders": [ { "user_id": 1486120, "full_name": "Nikhil Reddy", "title": "Founder/CEO", "founder_bio": "CEO @ Datoric. Prev Quant & SWE Intern. Math/Econ/CS @ UChicago; Bypassed Google OAuth 2.1 and QA systems on data annotation platforms to automate training tasks in 2023, then disclosed the loopholes to the affected platforms.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/nikhilrreddy", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/85cf42094155ddf41c9ca0e99b42a85e2ed763d1.jpg" }, { "user_id": 2310490, "full_name": "Jeffrey Lin", "title": "Founder/CTO", "founder_bio": "CTO @ Datoric. Prev AI/ML & SWE Intern. Math & CS & Robotics at NYU. Created a bounding box algorithm for more efficient labeling as Head of Data for the NYU Robotics Team. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/jeff-j-lin", "x_url": "https://x.com/Jeffjinlin", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/4f3068998c884af4a2668f4da4bbe3efa8e796d3.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 108251, "title": "Datoric | Trustworthy data for the next generation of models", "tagline": "Collecting the human experience AI needs, securely.", "body": "Datoric provides custom, security-first training data for voice models, robotics, and world models. Our private, project-specific contributor systems have led to a dramatic improvement in quality for customers and helped us generate nearly seven figures in revenue over the last 30 days.\n\n## **Our background**\n\nIn 2023, we built large-scale automations that completed paid LLM training tasks across major crowd-work platforms by bypassing QA testers. We earned six figures in three months, and even helped fix some of the vulnerabilities that we found. From this experience, we learned just how easily training data can be compromised, and how it can lead to unsafe model behavior.\n\n\n\n## **The problem**\n\nLabs are searching for the high-quality, authentic human data their models need, but quality and provenance become harder to control at scale, especially when relying on outside data providers. Poor-quality or compromised data can distort model behavior, while unqualified contributors, automated submissions, unclear instructions, and weak consent records can still slip through QA. At the same time, data providers hold concentrated stores of customer data, contributor identities, and payments that attract fraud and security threats.\n\n## **What we built**\n\nDatoric manages the full collection process for multilingual speech, egocentric video, and action-conditioned data. We collect this data through more than 300,000 active contributors who record conversations, everyday tasks, and other project-specific activities through our private apps. Because Datoric has no public marketplace or open upload system, every collection can run inside its own isolated environment, keeping customer projects separate and limiting unauthorized access. Each recording also remains linked to the verified contributor who created it and the consent they provided. This blocks automated or manipulated submissions and protects quality throughout the collection process rather than leaving it to final QA.\n\n## **Traction**\n\nWe’ve seen a major uptick in quality compared to other providers, and in the last 30 days, Datoric has generated nearly seven figures in revenue.\n\n## **Our ask**\n\nWe would love introductions to teams building voice models, robots, or world models that need difficult, custom human-generated data.\n\nSend us your hardest collection specification at [**founders@datoric.com**](mailto:founders@datoric.com), or visit [**datoric.com**](http://datoric.com).", "url": "https://www.ycombinator.com/launches/S9z-datoric-trustworthy-data-for-the-next-generation-of-models", "votes": 4, "created_at": "2026-07-29T16:22:25.416Z" } ] }, { "id": 31276, "name": "TensorTest", "slug": "tensortest", "former_names": [ "Ritivel", "ritivel", "Ritivel" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/93d28967db9fc99eca39bf1e1c109e9fdfea90d7.png", "website": "https://www.tensortest.com", "all_locations": "San Francisco, CA, USA", "long_description": "TensorTest (formerly Ritivel) builds evaluations, reinforcement-learning environments, and expert training data for AI labs developing creative agents, starting with video editing.\r\n\r\nAn agent can produce a valid video file without delivering an edit that meets the client’s brief. Labs need reliable ways to measure creative quality, identify failures, and give agents feedback they can learn from.\r\n\r\nOur first benchmark, TimelineBench, gives agents client video-editing jobs with raw footage, audio, and a brief. It evaluates the finished work using a verifier calibrated on thousands of blind judgments from professional editors. Across 16 agents we tested, the best passed only 26.8% of jobs. Most failures came down to creative craft rather than broken output.\r\n\r\nWe offer labs private evaluations to identify where their models fail, training environments that use our verifier’s scores as rewards, and expert-created training data, including professional editing demonstrations, preference pairs, and grading rubrics. These offerings connect evaluation to training: identify what an agent gets wrong, provide feedback on its attempts, and supply examples of better work.\r\n\r\nOur team comes from IIT Bombay and has trained language models and built agentic systems at Microsoft Research.", "one_liner": "Evals, RL environments and training data for creative tasks", "team_size": 3, "industry": "B2B", "subindustry": "B2B", "launched_at": 1768402832, "tags": [ "B2B", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2026", "status": "Active", "industries": [ "B2B" ], "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/tensortest", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2026/tensortest.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/tensortest", "x_url": "https://x.com/TensorTestTeam", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/93d28967db9fc99eca39bf1e1c109e9fdfea90d7.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/08801caf36f28c46fcbe90fb9027153da2f4d83b.png" ], "primary_partner": { "name": "Gustaf Alstromer", "url": "https://www.ycombinator.com/people/gustaf-alstromer" }, "founders": [ { "user_id": 939300, "full_name": "Pavan Kalyan", "title": "Founder/CEO", "founder_bio": "Cofounder and CEO @ TensorTest. \nPrev. @ Microsoft Research focused on training language models. \nGraduated from IIT Bombay with Btech + Mtech in Electrical+AI. \nPublished @ Neurips, ACL, Interspeech.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/tankalapavankalyan/", "x_url": "https://x.com/tpavankalyan23", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/c4b428e18d2e496d8d83d3de399dea8a25b3bd5c.jpg" }, { "user_id": 2779196, "full_name": "Nirmit Arora", "title": "Founder/CTO", "founder_bio": "Co-founder and CTO @ Ritivel. Prev @ Microsoft Research focused on making agentic systems safe. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/nirmitarora", "x_url": "https://x.com/NirmitArora26", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/3fc19dad7b2d51a113e38f3a00002d8716a82df6.jpg" }, { "user_id": 2979021, "full_name": "Gunin Gupta", "title": "Founder/COO", "founder_bio": "Cofounder and CBO @ Ritivel. \nPrev. @ AT Kearney focused on sustainability and energy process industries.\nGraduated with a bachelor's from IIT Bombay.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/guningupta/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/492637c609ab817e2febf7aa88286156cb2f269e.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 97327, "title": "Ritivel 💊 AI-native platform for Regulatory Document Submission", "tagline": "We help regulatory teams at pharma companies submit FDA documents 12x faster", "body": "Hey everyone 👋\n\nWe're Pavan (PK), Gunin and Nirmit – co-founders of Ritivel.\n\n## **TL;DR**\n\nRitivel is an AI-native workspace for regulatory and medical writing teams at life sciences companies. Our AI agents draft CTDs, CSRs, INDs, and BLAs in minutes instead of weeks, helping drugs reach patients faster.\n\n**Ask:** Can you introduce us to regulatory affairs leaders or CISOs at pharma companies? Contact us at [pavan@ritivel.com](mailto:pavan@ritivel.com)\n\n\n\n## **❌ The Problem**\n\nEvery month a drug is delayed from reaching the market costs pharma companies \\~$45M in lost revenue and more importantly, delays treatment for thousands of patients who need it.\n\nYet regulatory teams are stuck in a painful, months-long process to create FDA submissions like CTDs, CSRs, INDs, and BLAs:\n\n**1. Information overload** 📚 These submissions require synthesizing thousands of pages of clinical data, prior submissions, and regulatory guidance which is an overwhelming task even for experienced teams.\n\n**2. Document hunting & version chaos** 🔄 Teams waste hours digging through SharePoint folders, juggling multiple document versions, and chasing colleagues for the right files.\n\n**3. Endless follow-ups** 📧 Regulatory leads spend their days sending reminder emails, tracking down missing inputs, and managing status updates across teams.\n\n**4. Manual drafting bottlenecks** ⏳ Even with all the information gathered, the first draft alone takes weeks of painstaking work by highly specialized (and expensive) regulatory writers.\n\n---\n\n## **🔧 Our Solution**\n\nRitivel deploys AI agents that transform how regulatory teams work:\n\n**Draft the most common document types like CTDs, CSRs, INDs, and BLAs in minutes** ✍️ Our agents generate comprehensive first drafts directly in Microsoft Word, following FDA formatting requirements and pulling from your existing clinical data and prior submissions. Every data point includes citations linked back to the original source, so reviewers can verify with one click.\n\n![uploaded image](/media/?type=post&id=97327&key=user_uploads/939300/4d8b39f8-8230-4dcd-b654-c51b595d124e)\n\n**Intelligent document retrieval** 🔗 Connected to SharePoint and Veeva, our agents automatically source the right documents from across your organization: no more digging through folders or chasing down files.\n\n**Automated follow-ups** 📧 Integrated with Outlook, our agents handle reminder emails and follow-ups automatically, so you're not stuck chasing colleagues for missing documents.\n\n![uploaded image](/media/?type=post&id=97327&key=user_uploads/939300/aa5a53c8-8d14-4c11-a20c-fecd59724cae)\n\n**Accelerated project management** 🚀 By automating the document gathering and drafting loop, teams can focus on review and refinement rather than administrative overhead.\n\n\\\n**👬 The Team**\n\nWe're engineers who previously built AI copilots at Microsoft Research. We saw firsthand that we could create better, more specialized copilots for complex workflows.\n\nWe started by selling a generic copilot to process-heavy industries like life sciences to understand their regulatory workflows. After 50+ conversations with pharma professionals, we discovered the critical bottleneck in regulatory submissions and built Ritivel to solve it.\n\n![uploaded image](/media/?type=post&id=97327&key=user_uploads/939300/3447cd20-7180-467d-919f-e6ddb5839d7a)\n\n## **🙏 Our Ask**\n\n* **Know someone in pharma regulatory affairs?** We'd love an intro to regulatory affairs leaders or CISOs at pharma companies → [founders@ritivel.com](mailto:founders@ritivel.com)\n* **Work at a life sciences company?** Try Ritivel and see the difference →[ ](https://www.ritivel.com/)[www.ritivel.com](http://www.ritivel.com)\n\n---\n\n_Your intro could help get life-saving drugs to patients faster. Thank you!_", "url": "https://www.ycombinator.com/launches/PJn-ritivel-ai-native-platform-for-regulatory-document-submission", "votes": 172, "created_at": "2026-01-30T17:40:28.395Z" } ] }, { "id": 31300, "name": "Shofo", "slug": "shofo", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/cdc8fbfe28cb5c8e4405b92f33ce7b6536b88c17.png", "website": "https://shofo.ai/", "all_locations": "San Francisco, CA, USA", "long_description": "We are building the world’s largest video library. We've aggregated billions of videos into a searchable index and use agents to find and label the exact datasets a lab needs on demand.\r\n\r\nIf a lab needs 100K hours of cooking videos where someone is holding a pan, with reasoning annotations on top, our agents search the index, extract the matching subset, route it through our labeling pipeline, and deliver a custom dataset in days, not months.", "one_liner": "The World's Largest Video Library", "team_size": 4, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1768432806, "tags": [ "Data Labeling", "Infrastructure", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Winter 2026", "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/shofo", "api": "https://devasheeshg.github.io/yc-api/batches/winter-2026/shofo.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/shofoai/", "x_url": "https://x.com/shofoai", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/cdc8fbfe28cb5c8e4405b92f33ce7b6536b88c17.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": 2555175, "full_name": "Braiden Dishman", "title": "Founder/CEO", "founder_bio": "Co-Founder & CEO @ Shofo | UCSB Econ | Prev. AWS", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/braiden-dishman/", "x_url": "https://x.com/BraidenDishman", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/3e18cf46ac55fb35ab744fe0d1918aee373a3d85.jpg" }, { "user_id": 2980379, "full_name": "Andre Braga", "title": "Founder & Head of AI", "founder_bio": "Co-Founder & Head of AI @ Shofo | UCSB BS Statistics & Data Science | Prev. MIT", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/andre-braga-nlp", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/fd2a63604446d02437087711d56e433ed4cc9760.jpg" }, { "user_id": 3192733, "full_name": "Alexzendor Misra", "title": "Founder/CTO", "founder_bio": "Co-Founder & CTO @ Shofo | UCSB Dropout | Prev. CEO @ Correkt (43k users)", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/alexzendor-misra-15907a23b/", "x_url": "https://x.com/AlexzendorMisra", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/00c1c60b83a6b26bb5e20e358aefe74084124197.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 97200, "title": "Shofo - The World's Largest Video Library", "tagline": "Shofo is building the largest index of videos", "body": "### **TL;DR**\n\nShofo is building the largest index of videos.\n\n\n\n### **❌ The Problem**\n\nAI labs require massive, high-quality video data to train multimodal models, but clean, custom video datasets cost millions and take months to collect.\n\n### **✅ Our Solution**\n\nWe're building the world's largest library of videos by combining public videos collected from the open web and private videos aggregated from thousands of different sources. Both feed into a single, continuously updating index of billions of videos that we clean, search, and label to build custom datasets for AI teams.\n\nFor example, if a lab needs 100K hours of cooking videos featuring hand-object interactions with no scene changes, we query our index, extract the matching subset, and deliver a clean, structured dataset in days.\n\n### **🙏 Our Ask**\n\nIf you or anyone you know is working with computer vision, we'd love to chat. Reach out at [founders@shofo.ai](mailto:founders@shofo.ai) or fill out this form [here](https://www.shofo.ai/contact).\n\n### 📖 **Our Story**\n\nOur team met while building a previous startup called Correkt. Correkt was an AI search engine focused on multimodal content and reached over 40k users before pivoting to become Shofo.\n\n![uploaded image](/media/?type=post&id=97200&key=user_uploads/2355557/635c1bd9-0c05-4fcb-be47-39f092eec605)\n\n_From left to right: Braiden (COO), Alex (CTO), Bryan (CEO), and Andre (Head of AI)_\n\n### **😼 Appendix**\n\nWebsite: [https://www.shofo.ai](https://www.shofo.ai/)\n\nHugging Face: \n\nLinkedin: \n\nX: ", "url": "https://www.ycombinator.com/launches/PHk-shofo-the-world-s-largest-video-library", "votes": 45, "created_at": "2026-01-27T01:15:35.213Z" } ] }, { "id": 32752, "name": "Moving Atoms", "slug": "moving-atoms", "former_names": [ "Run Lobster", "RunLobster", "Demon AI", "RunDemon", "Avicenna Care", "Agent Insurance" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/ee4bd602221f77fc05dda940b7202523dd2b9d10.png", "website": "https://movingatoms.ai/", "all_locations": "San Francisco, CA, USA", "long_description": "Our first model, Atom 1, pre-trained on the internets video, and fine tuned for physics. \r\nIt has been action conditioned, so you can simulate robot actions within it.\r\nThis lets Robot Labs train their robots using GPUs instead of relying on human collected manual data that is slow and expensive.\r\nIn unseen and hard to replicate environments, this is the only path to getting ANY training data.", "one_liner": "Virtual Reality for Training Robots", "team_size": 2, "industry": "Industrials", "subindustry": "Industrials -> Manufacturing and Robotics", "launched_at": 1786481200, "tags": [ "Hard Tech", "Robotics", "Virtual Reality", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "status": "Active", "industries": [ "Industrials", "Manufacturing and Robotics" ], "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/moving-atoms", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/moving-atoms.json", "year_founded": 2026, "city": "San Francisco", "country": "US", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/ee4bd602221f77fc05dda940b7202523dd2b9d10.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Brad Flora", "url": "https://www.ycombinator.com/people/brad-flora" }, "founders": [ { "user_id": 1776350, "full_name": "Shehryar Saroya", "title": "Founder/CEO", "founder_bio": "Previously AI @ TikTok (Core Recomendation Algorithm). \nMaths @ Oxford.", "is_active": true, "linkedin_url": "https://linkedin.com/in/shehryar-saroya", "x_url": "https://x.com/SaroyaShehryar", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/eba8f18828b5f957663aea131e4a12ec89f9afbb.jpg" }, { "user_id": 2375000, "full_name": "Ahmad Saroya", "title": "Founder", "founder_bio": "Studied Mathematics @ Oxford, Trained Neural nets for GCHQ\nBritish Mathematical Olympiad Gold Medalist. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/ahmad-saroya-a05400237/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ab9fc0f9df4310e5df12d2db3ff69d64e603d611.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 32988, "name": "Markov", "slug": "markov", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/41a4a41e5482735389d2089a01ff2b862f56e307.png", "website": "https://www.markovstudios.com/", "all_locations": "San Francisco, CA, USA", "long_description": "Markov builds expert computer-use datasets for frontier AI labs. Our datasets include screen recordings + mouse/keyboard actions of professionals using CAD, design and enterprise software.\r\n\r\nWe work with the top AI labs and we’ve gotten more than 350k downloads on HuggingFace. Our long-term goal is to be the full stack platform for computer-use AI, starting with data and then the infrastructure to train and deploy computer-use models.\r\n\r\nDev and Harish are technical co-founders. Dev studied aerospace engineering at IIT Madras and was the youngest team member at Sarvam - India’s top AI lab. Harish is an Emergent Ventures grantee and his prior robotics work was featured in top national publications.", "one_liner": "Expert computer-use data for frontier AI labs", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Infrastructure", "launched_at": 1782277833, "tags": [ "Reinforcement Learning", "Data Labeling", "Data Engineering" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "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/markov", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/markov.json", "year_founded": 2026, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/markov-artificial-intelligence/", "x_url": "https://x.com/markov_labs", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/41a4a41e5482735389d2089a01ff2b862f56e307.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Harshita Arora", "url": "https://www.ycombinator.com/people/harshita-arora" }, "founders": [ { "user_id": 1783684, "full_name": "Dev Mandal", "title": "Founder/CEO", "founder_bio": "Co-Founder & CEO at Markov. Prev at Sarvam, Aerospace engineering at IIT Madras", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/dev-mandal-iitm/", "x_url": "https://x.com/DevvMandal", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/5a53c70cbcddfae2c6c21e32faac5865be5702b2.jpg" }, { "user_id": 2921049, "full_name": "Harish Ashok", "title": "Founder/CTO", "founder_bio": "Co-Founder at Markov. Previously founded Zenith, a robotics company building AI-assisted hardware workflow tools. ", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/harish-ashok", "x_url": "https://x.com/harishashok", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/fbfaea52b2a5bc846c87d04392062585acfb69ff.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 109874, "title": "Markov - Data for computer-use AI", "tagline": "150K+ downloads across our HuggingFace datasets", "body": "# \\*\\*\\*\\*\n\n**TLDR;**\n\n* **Markov builds the data infrastructure for computer-use AI**, collecting synchronized screen recordings, mouse/keyboard actions, and annotations from real workflows.\n* **We specialize in hard-to-source expert data** **across CAD, architecture, engineering**, and other professional applications.\n* **We’ve already sold 15,000+ hours of data to frontier AI labs** and open-sourced leading computer-use datasets with **150k+ downloads on HuggingFace**.\n\n\n\n**The Problem:**.\n\nWhile frontier models have gotten much better at reasoning, they still struggle to reliably complete long, multi-step workflows across real desktop software. **The biggest bottleneck is the lack of high-quality training data.**\n\nTraining these models requires **synchronized screen recordings, mouse and keyboard actions, and high-quality annotations from people doing real work**. This is especially scarce for expert workflows in domains like CAD, architecture, engineering, finance, and other professional software.\n\n**That’s where we come in:**\n\nWe work with a vetted network of expert contributors to collect high-quality computer-use data across:\n\n* Consumer applications\n* Productivity software\n* Accounting and finance\n* CAD, architecture, engineering, and other expert workflows\n\n![uploaded image](/media/?type=post&id=109874&key=user_uploads/1783684/ba673ff4-e489-4753-b11b-4797762834fb)\n\n**So far we've:**\n\n* Sold more than 15,000 hours of data to frontier AI labs\n* Open-sourced the [**largest dataset of computer-use screen recordings**](https://x.com/DevvMandal/status/2032104265121140924?s=20)(**150k+ downloads on HuggingFace**)\n* Released the [**largest open gaming computer-use dataset**](https://x.com/DevvMandal/status/2072357115847671990?s=20)\n\n**Our Ask:**\n\nIf you're building computer-use agents and need high-quality training data - or have interesting workflows you'd like collected - we'd love to chat **:)**\n\nBook a meeting with us at Or shoot us an email at [founders@markovstudios.com](mailto:founders@markovstudios.com)", "url": "https://www.ycombinator.com/launches/SaA-markov-data-for-computer-use-ai", "votes": 11, "created_at": "2026-08-11T18:00:00.277Z" } ] }, { "id": 33738, "name": "CoArena", "slug": "coarena", "former_names": [ "Coasty" ], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/1543623e6506eaf893b1c43e614688a6b2a32bda.png", "website": "http://coarena.ai", "all_locations": "San Francisco, CA, USA", "long_description": "CoArena is a live arena where anyone can use the world's top computer-use models racing two of them on the same real computer task and judging which one did it better. Every battle becomes something the AI labs can't build themselves, an honest test of their agents on real work, and the data to make them better.", "one_liner": "The biggest crowdsourced benchmark for Computer-Use", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1782294083, "tags": [ "AIOps", "Machine Learning", "Reinforcement Learning", "Data Labeling", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2026", "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/coarena", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/coarena.json", "year_founded": 2026, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/coastyai", "x_url": "https://x.com/coastyai", "fb_url": null, "cb_url": "https://www.crunchbase.com/organization/coasty", "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/1543623e6506eaf893b1c43e614688a6b2a32bda.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/cfec1f74a71f276cfe85bcc102647baf48c51fff.png" ], "primary_partner": { "name": "Brad Flora", "url": "https://www.ycombinator.com/people/brad-flora" }, "founders": [ { "user_id": 1194042, "full_name": "Prateek Jannu", "title": "Founder/CEO", "founder_bio": "Prateek Jannu, Co-founder & CEO - Stanford (MS. EE) and Purdue (BE. CE). Invented Solo-Synth-GAN, which generates labeled training data from a single example. Built Coasty to #1 on the OSWorld benchmark (82%), beating Agent S3, UiPath, Kimi K2.5, and Claude Sonnet 4.5. Now building CoArena.", "is_active": true, "linkedin_url": "https://linkedin.com/in/prateekjannu", "x_url": "https://x.com/prateekjannu", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/26ed45c60c34293a10373502e615a8c6499d356f.jpg" }, { "user_id": 2457255, "full_name": "Nitish Kovuru", "title": "Founder", "founder_bio": "Co-founder and CTO/COO of Coarena, launched Coasty prev, SOTA CUA framework on OSWorld at 83% accuracy. Did CS at Columbia (Vision track) and hold a BS in Computer Engineering from Purdue, where I TA'd a graduate-level AI course as an undergrad and later went on to build enterprise sales automation agentic systems for my previous company.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/nitish-kovuru-2423351b4", "x_url": "https://x.com/llmhubdev", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/29f588d1c332c655eee8ea69c5e1c27e9ec4fce9.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 111689, "title": "Coarena - The biggest crowdsourced benchmark for Computer-Use", "tagline": "We build the most advanced and biggest benchmark for computer use", "body": "We are officially launching CoArena (YC S26). Released 2 weeks ago, users have doubled every week since, and the arena went from $0 to $60,000 in revenue since. It's a live arena where the world's top Computer-Use models race to finish the same real computer task, and real people judge which one did it better, blind.\n\n\n\n![uploaded image](/media/?type=post&id=111689&key=user_uploads/1194042/43ddaa95-1f50-4b30-b0a4-9b085433ec3c)\n\n**The problem**\n\nEvery AI lab says their agent is the best at using a computer, and every claim points at a benchmark score. We know those scores well: earlier this year we built the #1 agent on OSWorld (82%). That's exactly how we learned the scores mean very little. Benchmarks are static, agents memorize them, and they say nothing about the messy tasks people actually need done.\n\n**What we built**\n\nOn CoArena, anyone can type a real task (\"find me the cheapest flight,\" \"assemble this expense report\"), watch two anonymous frontier agents race it live, and vote on the winner. Names are revealed only after you vote, so the judging is blind. Every battle becomes a fresh eval no model has ever seen, and the failure cases labs never catch: prompt injection, spending money when it should ask, silent side-effects.\n\n**What the arena has already shown**\n\n\\- Only \\~66% of agent runs finish their task at all 1 in 3 dies to a failure before any score exists\n\n\\- On identical tasks, the best frontier agent completes \\~87% of runs; the worst completes \\~42%\n\n\\- One major frontier model times out in roughly a third of its runs\n\n\\- In \\~7% of battles, human judges ruled that BOTH agents failed\n\n\\- The #1 spot on our leaderboard changed hands three times in three days, real tasks don't saturate\n\n**What you get**\n\nIt's completely free, including models you can't use anywhere else (Fable, Sol, and unreleased previews). Run your real errands with frontier agents, keep whatever they produce, and find out which model is actually best at YOUR kind of work, not the benchmark's.\n\n**Our asks**\n\n1\\. Use it, it's free. Bring your daily work: the flights, the spreadsheets, the research. Two frontier agents will race it, you keep the result, and your vote makes the leaderboard more honest: \n\n2\\. If you're an AI lab working on computer use or vision, or a robotics company building on vision, we'd love to talk. We can send you sample evals and data today: [founders@coasty.ai](mailto:founders@coasty.ai)", "url": "https://www.ycombinator.com/launches/T3R-coarena-the-biggest-crowdsourced-benchmark-for-computer-use", "votes": 59, "created_at": "2026-08-22T19:00:00.303Z" }, { "id": 105157, "title": "Coasty: The computer-use agent that doesn't break on real software", "tagline": "Coasty helps developers kill RPA.", "body": "**TL;DR: Coasty gives developers a computer-use agent that reliably automates back-office work in legacy desktop software where RPA scripts and other CUAs fail. It’s achieved 82.81% on OSWorld and it’s great at automating workflows where quiet mistakes cost the most: regulatory filings, patient records, payments, and gives developers precise control and observability over its work.**\n\n\\\n(_The portals and the data in it, are mockups_)\n\n**Replayable audit trail (real run):**  \n\n**OSWorld submission (Coasty scored 82.81% on OSWorld Verified, confirmed by the OSWorld team, with our latest internal run at 85.60%.):**[ https://github.com/coasty-ai/coasty-osworld](https://github.com/coasty-ai/coasty-osworld)\\\n\\\n**The Problem**\n\nMost back-office work doesn't happen in clean web apps. It happens in legacy desktop software, full of surprise pop-ups, wandering menus, and impenetrable UIs.\n\nToday, developers automating this back-office “real” work are stuck choosing between brittle RPA scripts and CUAs that demo beautifully and score high but break on contact.\n\nThe problem isn't getting an agent to click. It's trusting it when a silent mistake gets someone fired.\n\n### **Why Coasty is reliable where others aren't**\n\n**It recovers mid-task.** Coasty catches its errors and fixes them mid-run. If it’s crunching on a 20-step workflow and makes a misclick at step 3, it will recover and start from the error. First-pass accuracy is vanity. Recovery keeps a run alive.\n\n**It survives messy software.** Pop-ups, modals that steal focus, legacy desktop apps. Coasty doesn’t record clicks, it reads the screen. So when the UI changes overnight and every RPA script breaks, Coasty looks at the new screen and keeps going.\n\n**It knows when it's done.** An agent that thinks it finished but didn't is worse than one that fails loudly. Coasty verifies its own output and leaves a timestamped audit trail you can replay.\n\n**It swarms.** One task across forty state portals, each with its own ancient interface, all at once. Forty agents, one verified result set. What takes a team a week, is done overnight.\n\n### **Traction**\n\nCoasty has been verified by the OSWorld team and will be published soon, but also another internal benchmark we published (85.60%): < >.\n\nWe have paying customers on four continents, a few thousand active users, and we have actually grown 24x since Claude Cowork came out.\n\n### **Why us**\n\nWe started working on this idea when models weren’t good enough for computer use (last summer), so we’ve built frameworks to test and run the perfect combination of models and harnesses to see what works in practice. Prateek worked as an ML engineer at multiple startups and dropped his Stanford admit while I did CS at Columbia and built my previous company’s entire agentic sales automation. We’ve grown with this field and have been working on it even before it got much attention, so we understand where CUAs mess up and how to deal with these messes realistically.\n\n### **The ask**\n\nSend us one workflow your team runs by hand. We’ll tell you straight whether Coasty can run it reliably today. If it looks automatable, we’ll build a working proof of concept on your real software before a long sales process.\n\nWe’re especially looking for intros to operators and executives in accounting, tax, insurance, healthcare, compliance, and data privacy - anywhere teams are still babysitting RPAs, clicking through legacy portals, reconciling records, filing forms, or moving data across old desktop software.\n\nHere's our Calendly:[ https://cal.com/coasty/15min](https://cal.com/coasty/15min)", "url": "https://www.ycombinator.com/launches/RM5-coasty-the-computer-use-agent-that-doesn-t-break-on-real-software", "votes": 12, "created_at": "2026-07-10T17:30:25.886Z" } ] }, { "id": 36741, "name": "Perit.AI", "slug": "peritai", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/a35e9e2f6026717a2640d1e01b9215eaff2d4310.png", "website": "https://www.perit.ai/", "all_locations": "San Francisco, CA, USA", "long_description": "", "one_liner": "Expert-led data and RL platform for AI labs", "team_size": 47, "industry": "B2B", "subindustry": "B2B", "launched_at": 1789597716, "tags": [ "Data Labeling" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Fall 2026", "status": "Active", "industries": [ "B2B" ], "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/peritai", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2026/peritai.json", "year_founded": 2026, "city": "San Francisco", "country": "US", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/a35e9e2f6026717a2640d1e01b9215eaff2d4310.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Christopher Golda", "url": "https://www.ycombinator.com/people/christopher-golda" }, "founders": [ { "user_id": 1996169, "full_name": "Kshitij Dixit", "title": "Founder/CEO", "founder_bio": "Co-founder & CEO at Perit.AI. Previously co-founder/CEO at ZeoAuto (YC W20), scaled to millions of drivers across 140+ countries, $2M ARR profitable, 4.8 stars from 30k+ reviews. Led consumer payments at Jio Financial Services with 10M+ users. Former Senior Product Manager at OYO. IIT Kharagpur.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/kshitijdixit/", "x_url": "https://x.com/kshitij_dixit", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/d91115e1280ec1a6442f16ddbbb1457363eccab9.jpg" }, { "user_id": 3416920, "full_name": "Poorva Dixit", "title": "Founder/CTO", "founder_bio": "Co-founder & CTO at Perit.AI. Previously Software Engineer at Google, Atlassian and Uber. Co-founded SalaryBook. Dual degree in Computer Science from IIT Kharagpur.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/poorvadixit/", "x_url": "https://x.com/PoorvaDixit3", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/7557e5ae918e8a0a80d40a998b247782f7a30d4e.jpg" } ], "jobs": [], "news": [], "launches": [] } ]