Archestra.AI Announces $10M Seed 🎉 | Blog | Archestra<br>Star us on GitHub ⭐<br>Star us on GitHub ⭐

Written by<br>Matvey Kukuy, CEO

Today I'm happy to share that we've raised a $10 million Seed round, led by 20VC . That brings our total funding to $13.5M. But the number I'm prouder of is a different one: multiple Fortune 500 companies are now running Archestra in production, connecting AI agents to real corporate data .

Who's backing us

20VC (Harry Stebbings) leads the round. Joining are 20 Product (Will Wu, CPO at Match Group), Olivier Pomel (CEO and co-founder of Datadog), Visible Ventures with Carolyn Everson (Senior Advisor at Permira, Board Member at Coca-Cola and Disney), Tenacity Capital (AJ Tennant, ex-VP Sales at Slack and Glean), Kieran Flanagan (CMO at HubSpot), Insiders (co-founders of ESL FACEIT Group), Debo Dutta (Chief AI Officer at Nutanix), Alexandre Berriche (Founder of Fleet), Gerhard Eschelbeck (former CISO at Google), Mark Goldberger (VP Sales at Metaview), and Declan Kelly (Founder & GP at Foreword).

Our pre-seed investors — Concept Ventures, Zero Prime Ventures, Celero Ventures, and Aloniq — came back in.

"Founders fail mostly for two reasons at seed. Either they quit because they lack the grit. Or they lack the ability to build a product. Here we have psychopaths who will not quit. Plus, they are top 0.1% engineers."

Harry Stebbings, 20VC

I'm taking "psychopaths" as a compliment 🤭

What changed in ten months

When we wrote the pre-seed post, we said enterprises would push hard on AI agents and immediately run into security and compliance walls. That's exactly what happened — just faster and more painfully than I expected.

The first wave of AI pilots has already wrapped up across the Fortune 500, and a lot of teams are arriving at the same uncomfortable conclusion: scaling agents isn't a model problem. It's an infrastructure problem. You can have the best model in the world, but if you can't connect it to your data without risking exfiltration, you've got an expensive demo, not a deployment.

That's the gap we've been filling. Archestra is a horizontal platform that lets in-house AI teams open up data access to business users without the usual nightmare scenarios. It plugs into the AI tools you already use, deploys in days, and gives you deterministic guardrails, governance, and observability out of the box.

Until now, teams have had to choose between locking agents down so hard they're useless, or letting them loose and praying. We don't think that's a real choice. Our platform guarantees 100% protection against data leaks while actually accelerating agent development. Both things, at once. That's the whole bet.

The other thing I underestimated was how fast the open-source community would show up:

3,800+ GitHub stars

1,000+ community members

57 contributors

To everyone who filed an issue, sent a PR, joined Slack, or just starred the repo — thank you. You're the reason this works.

What we'll do with the money

We're expanding GTM and engineering, deepening the open-source ecosystem, and pushing harder into security-conscious industries where we already have pull (legal, procurement, ops, regulated finance).

If you're an engineer who wants to build the security layer for the agentic enterprise, or a GTM operator who's tired of selling AI features that don't actually ship — we're hiring. Come build with us.

And if you're on an AI or Platform Team

If you're wrapping your head around the MCP, CLI, API, auth, observability, cost-management, and security knot of the modern enterprise AI stack — book a demo, or just spin it up, it's open source.

Star us on GitHub · Join our Slack

Press: margaret@archestra.ai