Plurai raises $13m to test AI agents before deployment
Plurai's simulation platform recreates an agent's working environment to expose failures and improve its models.

The difficult part of an AI agent is often the situation its builder never thought to test.
What happened
Plurai announced $13 million in seed funding led by Team8, with U&I, Velocity and Mercer participating, and opened its platform to developers.
Its software recreates an agent's working environment from a customer's tools, knowledge and policies. Simulated users then exercise different scenarios. The resulting data supports fixes and the training of smaller models for specific tasks.
Why it matters
An agent can follow the happy path while struggling with conflicting instructions, missing information or unusual requests. Simulating those interactions gives teams a way to examine failures before customers encounter them.
The bigger picture
Agent development creates demand for infrastructure beyond the underlying model. Testing, evaluation and ongoing supervision become recurring engineering work as agents gain access to business systems.
