about

In 2018, I co-founded Poker Scientist with a mission to democratize game theory for poker players. This taught me more about building production software than any CS degree could.

The real "aha moment" came when I realized that great engineering isn't about the latest framework—it's about solving hard problems for real people. That insight led me from frontend dashboards to Nash equilibrium solvers, and eventually to AI systems that augment human decision-making.

Today, I'm building what I call the "AI Agent Ecosystem" — four interconnected projects that form the infrastructure for AI agents that need identity, memory, and knowledge. It's production code running daily, not demos or proofs of concept.

Where others see complexity, I see patterns. Where others context-switch, I deep-dive for days. The result: coherent architectures, not patched prototypes — that's how I built a Nash equilibrium solver and shipped 4 production AI systems in parallel.

What drives me: building tools that make complex decisions easier. I'm happiest when surrounded by hard problems and clear ownership.

I work with teams shipping production AI — Claude Code architecture, MCP server development, RAG pipelines, and multi-agent orchestration. Based in Berlin, working globally.