Sycamore Raises $65M Seed to Build the Operating System for Enterprise AI Agents

There’s a tension building inside the enterprise right now. On one side, the promise of AI agents moving faster than any team ever could. On the other, the quiet reality that one wrong move from that same agent can ripple through a system like a bad trade on margin. Speed is easy to sell. Trust is harder to earn. That’s the space Sycamore stepped into, not with noise, but with structure.

Sycamore, based out of Palo Alto, just pulled in a $65M seed round. Coatue and Lightspeed Venture Partners co-led the charge, with Abstract Ventures, Dell Technologies Capital, 8VC, Fellows Fund, and E14 Fund all leaning in. That’s not a casual cap table. That’s a room full of people who’ve seen enough cycles to know when something foundational is forming.

Credit where it’s due, Sri Viswanath, Founder and CEO, didn’t come into this guessing. Former Atlassian CTO. Decades in enterprise systems. The kind of background where you don’t just build tools, you understand the consequences of them running at scale. And that matters, because Sycamore isn’t building another shiny AI layer. They’re building what they call an agent operating system. Not a feature. Not a plugin. An operating system. Let that sit for a second.

We’ve spent years teaching machines how to think. Now the real question is how to let them act without burning the house down. Sycamore’s answer is structured control. Discovery, deployment, observability, governance. Every agent accounted for, every action auditable, every permission earned. Less “move fast and break things,” more “move precisely and keep your job.”

And here’s where it gets interesting. The system doesn’t just run agents. It builds them, evolves them, and learns from them. Natural language in, production-ready systems out. Agents that don’t just execute tasks, but accumulate institutional knowledge over time. That’s not automation. That’s compounding intelligence. They’re already working with Fortune 500 companies, which tells you this isn’t theory dressed up as a demo. It’s being tested where the stakes are high and the tolerance for nonsense is low.

The takeaway isn’t just that Sycamore raised a big seed. Plenty of companies do that. The signal is where the money went. Into infrastructure, governance, and trust. The unsexy layers that actually determine whether AI sticks inside an enterprise or gets quietly shelved after the pilot. Everyone wants autonomous agents. Very few are building the rails to keep them accountable.

Sycamore is betting that the future of AI in the enterprise won’t be defined by how smart the agents are, but by how well they’re controlled. And judging by who showed up with capital, that bet is starting to look less like a theory and more like the next operating layer quietly taking shape underneath everything.