Patronus AI raises $50M Series B to build simulated environments for testing AI agents

Patronus AI, a San Francisco-based startup founded in 2023 by former Meta AI researchers Anand Kannappan and Rebecca Qian, announced a $50 million Series B funding round led by Greenfield Partners. Notable Capital, Lightspeed, Datadog, and Samsung also participated in the round, bringing Patronus’s total funding to $70 million.

The company builds simulated ‘digital world models’—replicas of websites and internal systems—that are used to stress-test AI agents’ performance across complex, real-world scenarios. Unlike traditional benchmarks, these environments allow agents to attempt tasks iteratively, with reinforcement learning rewarding successful completions and penalizing errors.

According to Glenn Solomon, a managing director at Notable Capital, demand for Patronus’s simulated environments is “nearly insatiable,” with virtually every frontier AI lab and many emerging startups now as customers. Patronus’s revenue has grown 15-fold over the past year, fueling investor interest in the round.

Patronus’s approach is designed to catch behaviors that benchmarks might miss, such as agents taking shortcuts that lead to incorrect task completion. The company compares its method to Waymo’s use of synthetic worlds to test autonomous vehicles against rare hazards like severe weather or a child running into the street.

While Patronus currently focuses on verifiable domains such as software engineering and finance, the company plans to expand into areas that are harder to verify. Anand Kannappan, co-founder and CEO, noted that the goal is to create environments where agents can operate for extended durations—such as 10 hours, 10 days, or even 10 weeks—without human oversight.