CRV Leads Raindrop Series A, Total Funding Hits $50M
Fundraise Insider tracks newly funded startups each week and delivers verified sales leads of C-suite contacts straight to your inbox, so you reach the right people while the funding is still fresh. See pricing.
Raindrop, a San Francisco company that monitors AI agents after they go live, has closed a funding round led by CRV. The Series A takes total funding to $50 million, according to the company’s announcement on September 17. Raindrop did not disclose the size of the round itself.
Lightspeed Venture Partners and Y Combinator, both existing investors, joined the round. Lead researchers from OpenAI, Anthropic, and Thinking Machines invested personally. The California company also counts Figma Ventures and Vercel Ventures among its backers.
The product reads production traces from AI agents and flags what Raindrop calls semantic anomalies. That covers hallucinated answers, misuse of tools, and the behavior shifts that arrive with a model upgrade. Engineering teams can see what changed, when it started, and which users were affected.
Alongside the funding, Raindrop released a product called Simulations in research preview. It replays real production traffic and a team’s existing test cases against a proposed change to an agent, then runs anomaly detection over the results. The argument is that conventional evals depend on test cases written ahead of time, so they mostly catch the failures a team already imagined. Simulations runs on every pull request and is meant to surface the behavior changes nobody predicted.
Raindrop pointed to METR research showing that the length of tasks agents can complete on their own roughly doubles every seven months, and a single run can now stretch over days and thousands of tool calls. The company also argues that Simulations gives teams outside the frontier labs the same kind of testing process those labs use internally, citing OpenAI’s work on deployment simulation and Anthropic’s synthetic universes for stress testing agents.
Customers include Vercel, Framer, and Clay, along with unnamed Fortune 100 enterprises in healthcare and logistics.
CEO Zubin Koticha founded Raindrop with Ben Hylak and Alexis Gauba. The team includes engineers who built fraud detection models at Robinhood and anomaly detection at Square, and security engineers from Segment, Semgrep, and Socket.dev have since joined.
The category is getting crowded. groundcover raised $100 million in July for observability in the AI era, Scaled Cognition took $100 million from Khosla in June for reliable agents, and Harvey bought Guardrails AI this month. Raindrop’s distinction is where it intervenes: it treats agent reliability as a detection problem rather than a design one, and it bets the pull request is the moment that matters. With Simulations still in preview, that bet is not yet settled.
Software vendors selling developer tools and infrastructure can receive verified contacts at newly funded software startups every week through B2B SaaS Leads.