Raindrop Raises $35M to Monitor AI Agents

AI agents create a new monitoring problem: software can be technically online while still making the wrong decision. Raindrop has raised $35 million to build observability for that new class of systems.

What happened

Raindrop closed a $35 million Series A led by CRV.

The platform is designed to monitor autonomous agents and help companies understand how they behave across longer workflows, including when they take unexpected actions or produce poor outcomes.

Traditional observability tools focus on infrastructure health, latency and software errors. Raindrop is extending that concept toward agent behaviour.

Why it matters

As agents gain permission to access company systems, call APIs and take actions, failures become harder to diagnose than a conventional software crash.

Companies need to know not only whether an agent completed a task, but how it got there, which tools it used and where its behaviour diverged from expectations.

The bigger picture

A new AI operations stack is forming around evaluation, monitoring, permissions and auditability.

Raindrop sits inside that layer. If autonomous agents become normal enterprise infrastructure, monitoring their decisions could become as standard as monitoring servers and applications today.