The archive · AI & Models · Strategic decision · 2025–2026
Sycamore raises $65M seed to become the operating system for enterprise AI agents
Ex-Coatue partner and Atlassian CTO Sri Viswanath raised $65M to build a trusted OS for deploying fleets of enterprise AI agents safely.
Sycamore
What the business is
A Palo Alto startup building a 'trusted agent operating system': a platform to discover, build, deploy and observe fleets of enterprise AI agents with security, governance and human control.
Starting capital：$65M seed led by Coatue and Lightspeed (2026-03-30)
How it started
Sri Viswanath left his Coatue investing role in fall 2025 after 20+ years building enterprise platforms at Sun Microsystems, VMware and Groupon, and as Atlassian CTO scaling an engineering org past 7,000. He founded Sycamore to build the agentic orchestration layer; the $65M seed came together through long-standing relationships.
What happened
Sycamore's platform lets employees describe tasks in natural language; agents build the needed applications and integrations, start heavily monitored, and earn more autonomy as they demonstrate reliability. The startup said it had traction with large enterprise customers but declined to name them, while entering a crowded race that includes OpenAI's Frontier, Anthropic's Cowork, Microsoft Foundry, AWS Bedrock AgentCore and startups like Isara ($94M), Airia and Port ($100M each).
How it ended up
Still live and early: as of 2026-03-30 Sycamore had $65M in the bank and was scaling engineering and applied-AI teams to move agents from lab to production, betting its governance-first positioning and veteran founder would win the category.
Background
Sycamore is a Palo Alto startup founded by Sri Viswanath, a former Coatue investor and ex-CTO of Atlassian, betting that enterprises will adopt AI agents at scale only through a governed operating system rather than a single tool. On March 30, 2026 it announced a $65M seed round led by Coatue and Lightspeed, with angels including former OpenAI chief scientist Bob McGrew, Intel CEO Lip-Bu Tan, Databricks CEO Ali Ghodsi and researcher François Chollet.
The product thesis is that agents work fine in demos but are too risky to let loose inside companies without infrastructure for safety at scale. Sycamore's platform lets employees describe a task in natural language; agents create the necessary applications and integrations, then start heavily monitored and 'earn' more autonomy as they prove reliable. Viswanath's contrast with rivals: most tools layer agents onto existing workflows, while Sycamore starts from the problem and rebuilds the flow with agents, back-end systems and data integrations.
The bet is on category timing as much as technology. Viswanath argues the next generation of enterprise software will be autonomous and adaptive, and that nobody yet owns the trust and governance layer — but Sycamore faces OpenAI's Frontier, Anthropic's Cowork, Microsoft Foundry, AWS Bedrock AgentCore and well-funded startups like Isara, Airia and Port. Sycamore said it had traction with large enterprise customers but declined to name them as of the round.
What has to be true
- Enterprises hesitate to let agents loose without infrastructure for safe deployment at scale; Sycamore makes governance the product rather than an add-on.
- The founder is the pitch: 20+ years running enterprise platforms at Sun, VMware, Groupon and Atlassian gave Viswanath credibility with investors where younger founders lacked it.
- The market is widely expected to be huge — model makers and cloud providers are all racing for it — but no one owns the trust layer yet.
- The tiered trust model turns a safety problem into a feature: agents earn autonomy with evidence, which is how risk-averse enterprises buy.
What can be applied
A mega-seed can rest on operator pedigree: 20 years running enterprise platforms raised $65M on relationships before product proof, betting governance, not model quality, would be the moat.
Aftermath
As of March 30, 2026, Sycamore had announced its $65M seed and was using the money to scale engineering and applied-AI teams and move its agents from lab into production, focused on trust architectures and multi-agent coordination. It had not named customers or disclosed revenue; its stated goal was to become the foundational operating system for autonomous enterprise AI, competing with model makers and cloud providers that want the same layer.
Sources
- Former Coatue partner raises huge $65M seed for enterprise AI agent startup
- Sycamore raises $65M from Silicon Valley heavyweights to build governance layer for enterprise AI agents
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card