Patronus AI raises $50M Series B to stress-test AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Patronus AI raised a $50 million Series B led by Greenfield Partners, with Notable Capital, Lightspeed, Datadog, and Samsung participating, bringing total funding to $70 million.
- Anand Kannappan and Rebecca Qian founded Patronus in 2023 after working as researchers at Meta AI, building "digital world models" that replicate websites and internal systems to stress-test agents after reinforcement learning.
- Glenn Solomon, managing director at Notable Capital, said virtually every frontier AI lab and many emerging startups are now Patronus customers, with revenue growing 15-fold over the past year.
- Kannappan compared the company's approach to Waymo's synthetic-world training for autonomous cars, noting agents tend to "take shortcuts" — a failure mode Patronus is designed to catch and hold models accountable for.
- The platform currently focuses on software engineering and finance, with Kannappan saying the next step is environments where agents can run continuously for 10 hours, 10 days, or 10 weeks.
- Human-data firms like Mercor and Surge help with reinforcement learning, but Patronus says its edge is evaluating agent behavior without human involvement; its primary competition is the in-house evaluation teams AI labs already run.
Why it matters: With 15-fold revenue growth and adoption described as 'nearly insatiable' across frontier AI labs, Patronus is positioning itself as external verification infrastructure for the agent economy — a third-party check on whether autonomous systems actually complete the tasks they were built for before users trust them with financial analysis or trip booking.
