Patronus raises $50M for AI agent testing
Patronus AI raised a $50M Series B to build simulated environments for testing AI agents in complex workflows.
AI agents are moving from demos into real business workflows, and reliability is becoming a serious bottleneck. Patronus AI is building the testing layer companies need before trusting agents with higher-stakes work.
What happened
Patronus AI raised a $50M Series B led by Greenfield Partners, with participation from Notable Capital, Lightspeed, Datadog and Samsung.
The startup builds simulated digital environments to test how AI agents perform across complex scenarios, especially in software engineering and finance workflows. The round brings total funding to $70M, and the company says revenue grew 15x over the past year.
Why it matters
This is a strong AI agent infrastructure signal.
As companies deploy agents into real workflows, they need more than benchmark scores. They need to know whether agents can complete tasks reliably, handle edge cases and avoid costly mistakes.
The bigger picture
The AI agent market will need a trust layer. Evaluation, stress-testing and reliability tools could become core infrastructure as agentic systems move into production.