Raindrop said a new Series A brings its total funding to $50 million, led by CRV, and introduced Raindrop Simulations, software meant to surface AI-agent failures before changes ship.
In a Sept. 17 company post, Raindrop said Simulations replays production traffic and existing tests against proposed agent-harness changes on every pull request, then applies its anomaly detection to flag unexpected behavior. The company contrasted that with traditional evals that, it argues, mostly catch failures teams already expect. Raindrop said it developed Simulations with large enterprise partners and is widening early access, with broader availability planned in the coming month.
Raindrop’s announcement names CRV as lead and says researchers and executives at Anthropic, OpenAI and Thinking Machines participated, without stating the Series A check size or valuation. Axios Pro separately reported the round at $35 million with Lightspeed and Y Combinator participating—figures Raindrop’s own post does not confirm. Co-founders framed the raise as backing for tools that catch agent failures “big and small” as buyers watch a wave of agent-security headlines from frontier labs.