Siavash Ghorbani and Kaj Drobin have been partners for many years, first as conveners of Stockholm’s innovation ecosystem, and later as co-founders of Tictail. After Tictail was acquired by Shopify, they built Shop and Shop Pay to serve millions of merchants and hundreds of millions of consumers, learning what it takes to build products at true hyperscale. They know that velocity is the definitive competitive advantage for product development and that coordination overhead is often the biggest impediment to speed. That’s why they’re building Stilla, the first multiplayer AI agent for product teams.

Stilla reimagines collaborative product management with intelligence from the ground up, designed not just for individual productivity but for cross-team and multi-tool alignment. We believe they are creating new infrastructure for AI-native product workflows and are proud to partner with Siavash, Kaj, and the Stilla team from day one.

‍Product Development for the AI Era

Ask any product leader about their biggest bottleneck and you’ll hear a familiar refrain: teams moving fast in different directions, lengthy call notes where critical decisions get lost, GitHub pull requests that contradict the product specifications. Despite everyone working hard, effort often fails to translate into meaningful progress.

Part of the challenge is the very tools meant to accelerate product development. Product managers spend their days translating between a multitude of tools, chasing down context across different platforms, and running alignment meetings just to keep everyone synchronized. AI-native workflows are only making this challenge harder, with agents duplicating work, contradicting each other, and optimizing for conflicting goals.

‍Unlocking Speed Through Coordination

Stilla transforms the nature of AI-enabled product development collaboration. It acts as a shared brain across the tools that product teams actually use—Slack, Linear, GitHub, Notion, and more. Their platform captures context, connects the right information across teams, and orchestrates actions in real time. The result is a single source of truth for the entire team.

This multiplayer architecture forms the foundation for agentic coordination: collaborative intelligence that doesn’t just watch, but acts to maintain alignment, accelerate development with speed, and empower teams to ship great products faster.

‍We sat down with Siavash and Kaj to discuss what they’re building at Stilla. This interview has been edited for length and clarity.

‍What insight are you building on that is obvious to you but not to others?

We believed AI would make individuals 10-100x faster, but alignment wouldn't keep up. When everyone moves at that speed, you can't coordinate through weekly syncs and status meetings, especially when everyone is building with “islands of intelligence”: your Cursor context, their Claude projects, someone else’s Figma AI. The bottleneck has shifted from capability to coordination—and the fix is automating alignment itself, at the pace of AI.

How are you approaching early customer discovery?

Hands-on design partners. We tested with companies at different sizes like Spotify, Ramp, Lovable, Polar, and Legora throughout 2025 before launching. One thing became clear: while entry points may differ, product organizations get it. 

As founders, how do you think about perseverance and adaptation?

We expected models to get so good that the entire landscape of coding editors would be disrupted. We bet on a future where AI agents could independently carry out long-running, complex tasks. So we built toward that: a system that helps humans stay aligned. And if it can align humans, it can align autonomous AI agents. The product worked 50% of the time when we started in 2025. By year-end, models were ready and we decided it was time to launch.

What impact do you want this company to have in the world?

Eliminate the work around the work. Status updates, context-sharing, re-explaining decisions, chasing follow-ups—all of it. With coordination handled automatically, teams can build products that were previously impossible, not because of technical limitations but because collaboration cost was too high.