Raindrop builds monitoring software for engineering teams running AI agents in production. The company has raised a $35 million Series A led by CRV, co-founders Zubin Koticha and Ben Hylak told Axios Pro. Its platform captures an agent’s messages, tool calls, retries and errors so teams can find failures that do not necessarily surface as conventional software crashes.
Raindrop says the workflow starts by logging each production run in one place. Its software reads those traces for problems such as hallucinations, repeated loops and broken tools, then sends alerts into Slack, where engineers can investigate an issue, ask questions and define additional behaviors to monitor.
After engineers ship a change, they can place it behind a feature flag and test it against live traffic to see whether the problem returns. Raindrop also offers an open-source local workshop for inspecting traces, replaying runs and debugging before an agent reaches production, connecting development work with continuing monitoring.
Raindrop says engineering teams at Speak, Vercel, Clay, Framer, AngelList and Tolan use its software. It sells the hosted product through a free tier, a usage-priced Pro subscription and custom enterprise plans that add controls such as audit logs, single sign-on and data exports.
Axios described the financing as a $35 million Series A led by CRV and attributed the information directly to Koticha and Hylak. The new round follows $15 million in funding that Raindrop announced in December 2025, led by Lightspeed Venture Partners.
The business now spans local debugging, production monitoring and enterprise controls. Its next test is whether automatic detection and remediation can remain dependable across the varied agents, tools and failure modes that customers operate.