Research and engineering under one roof
Patronus AI is a behavioral research lab for AI. We study how frontier models behave and investigate where they fail. We turn what we find into the training data that moves model capabilities forward.
Mission
Our mission is to deeply understand model failures. Not the cases where a model gets the answer wrong, but the ones where it technically finishes the task and does something no one wanted along the way. Tests rarely catch this type of failure. We build environments that surface them, then work out why they happen.
Founders
Patronus AI was founded in 2023 by Anand Kannappan and Rebecca Qian. Our researchers and engineers come from Meta Superintelligence Labs, Amazon AGI, Apple and Google.
Anand Kannappan
Co-founder & CEO
“Reliability for AI agents will require the equivalent of ‘five nines’ for software systems. The path to getting there runs directly through simulation.”
Rebecca Qian
Co-founder & CTO
“The next frontier of AI depends on scalable simulation and trustworthy evaluation. We’re building the infrastructure to make that possible.”
Backing
September 2023
$3M Seed
Led by Lightspeed Venture Partners, with participation from Factorial Capital and angel investors including Amjad Masad and Gokul Rajaram.
Read the Seed announcement →May 2024
$17M Series A
Led by Notable Capital, with participation from Lightspeed Venture Partners, Datadog, Factorial Capital and angel investors.
Read the Series A announcement →June 2026
$50M Series B
Led by Greenfield Partners, with participation from Lightspeed Venture Partners, Notable Capital, Datadog, Samsung, Gokul Rajaram, Factorial Capital, and a cohort of AI leaders across the labs and neolabs.
Read the Series B announcement →
Investors
- Gokul Rajaram
- Amjad Masad
Explore the lab
- Domains →
The areas where we help push the frontier.
- Research →
Papers and benchmarks from the lab.
- Blog →
Write-ups of what we learn.
- Data →
License your data to Patronus AI for research and training.