Sycamore Raises $65M Seed to Build the Agent OS
The biggest seed round in the agentic space just dropped. Sycamore, founded by Sri Viswanath — former CTO of Atlassian and more recently a partner at Coatue — pulled in $65 million led by Coatue and Lightspeed. The angel list reads like a tech industry power directory: Bob McGrew (former OpenAI Chief Scientist), Lip-Bu Tan (Intel CEO), Ali Ghodsi (Databricks CEO), BJ Jenkins (Palo Alto Networks President), François Chollet, and Frederic Kerrest (Okta co-founder). Dell Technologies Capital, Abstract Ventures, and 8VC also participated. That's not just money. That's a signal.
What is Sycamore actually building? An operating system for enterprise AI agents. Not another agent framework — a governance layer. The pitch is simple: companies can build individual agents just fine, but the moment you need fifty agents running simultaneously, things fall apart. No central identity management, no policy enforcement, no way to prevent agents from stepping on each other's toes. Sycamore calls this "operational gravity" — the invisible force that keeps agent deployments stuck in pilot mode.
The platform has a trust-based autonomy system where agents earn more freedom as they prove reliability. Think of it like employee onboarding — new agents start supervised, and gradually gain access to more critical systems. Add natural language task definition, cross-agent coordination, and institutional learning where agents capture organizational knowledge over time.
This is the Meta Dreamer thesis (acqui-hired for $56M just weeks ago) validated at scale. The "Agent OS" category is becoming real, and the fact that both investors and operators are piling in tells you something. The real question isn't whether enterprises need agent governance — it's who builds the one that actually works.
https://techcrunch.com/2026/03/30/former-coatue-partner-raises-huge-65m-seed-for-enterprise-ai-agent-startup/
← Back to all articles
What is Sycamore actually building? An operating system for enterprise AI agents. Not another agent framework — a governance layer. The pitch is simple: companies can build individual agents just fine, but the moment you need fifty agents running simultaneously, things fall apart. No central identity management, no policy enforcement, no way to prevent agents from stepping on each other's toes. Sycamore calls this "operational gravity" — the invisible force that keeps agent deployments stuck in pilot mode.
The platform has a trust-based autonomy system where agents earn more freedom as they prove reliability. Think of it like employee onboarding — new agents start supervised, and gradually gain access to more critical systems. Add natural language task definition, cross-agent coordination, and institutional learning where agents capture organizational knowledge over time.
This is the Meta Dreamer thesis (acqui-hired for $56M just weeks ago) validated at scale. The "Agent OS" category is becoming real, and the fact that both investors and operators are piling in tells you something. The real question isn't whether enterprises need agent governance — it's who builds the one that actually works.
https://techcrunch.com/2026/03/30/former-coatue-partner-raises-huge-65m-seed-for-enterprise-ai-agent-startup/
Comments