Patronus AI, Inc.

Based in California, US, and founded in 2023, Patronus AI, Inc. is a behavioral AI research lab focused on developing training environments and simulation infrastructure for advanced AI systems. The company conducts research on AI behavior, reliability, and reinforcement learning, creating realistic digital environments that enable AI agents to learn, adapt, and be evaluated in complex workflows. The company was co-founded by Anand Kannappan (CEO) and Rebecca Qian (CTO).

In June 2026, Patronus AI, Inc. raised USD 50 million in series B funding led by Greenfield Partners, including Samsung NEXT Ventures, Factorial Capital, Notable Capital, Lightspeed Venture Partners, and Datadog, Inc. The series B funding values Patronus AI, Inc. at USD 406.

43 million post-money. Patronus AI provides reinforcement learning environments and digital world models designed to support the training and development of AI agents. Its reinforcement learning environments create domain-specific settings where agents can perform tasks, interact with simulated systems, and learn through feedback and reward signals. Its digital world models simulate software environments and predict how those environments respond to agent actions, allowing agents to train without relying solely on live systems.

Together, these offerings support the development, training, testing, and improvement of AI agents across complex digital tasks and workflows, providing controlled environments for building agents that can operate within specialized software systems and real-world scenarios. The company generates revenue by providing AI evaluation, monitoring, and optimization software for organizations developing and deploying AI systems and agents. Its revenue streams are derived from its AI evaluation platform, agent evaluation and analysis tools, monitoring and analytics capabilities, Digital World Models, generative simulation environments, benchmarking solutions, and dataset generation services. The company plans to use the June 2026 funding to expand its research and engineering teams and invest in compute and infrastructure to develop and scale its digital world models for testing AI agents.