Braintrust
AI observability platform that monitors and evaluates AI models in production.
- Founded
- 2023
- Headquarters
- San Francisco, CA
- Latest Round
- Series B
- Est. Valuation
- ~$800M
Investment Thesis
Braintrust provides AI-focused observability and evaluation for monitoring AI models in production. Enables teams to run experiments, track performance, catch regressions, and debug issues.
Customers include Notion, Replit, Cloudflare, Ramp, and Dropbox. Reached $800M valuation with $80M Series B.
There's no reliable way to test whether AI prompts and models actually give the right answers in production, and failures are invisible until customers complain.
A complete AI observability toolkit for testing, logging, evaluating, and improving AI models in production โ catching regressions and hallucinations before users do.
As AI agents make more autonomous decisions in critical workflows, rigorous testing and monitoring isn't optional โ it's the difference between a demo and a product that enterprises trust.
Used by Notion, Stripe, Replit, Cloudflare, Ramp, and Dropbox; reached $800M valuation with $116M raised, becoming the standard for AI-native companies shipping to production.
Ankur Goyal (CEO) combines deep product experience at Figma and Brex with systems engineering expertise, building the developer tool that AI teams didn't know they desperately needed.