Patronus AI, founded by former Meta AI researchers, has raised a $50M Series B to build simulated 'digital world models' that stress-test AI agents. These environments replicate websites and internal systems, allowing agents to be evaluated and fine-tuned via reinforcement learning across complex, real-world scenarios. Revenue has grown 15x in the past year, with virtually every frontier AI lab among its customers. The company currently focuses on verifiable domains like software engineering and finance, and differentiates itself by evaluating agent behavior without human involvement.
361 Impressions