Building at the edge of what AI systems can actually do.
- multi-agent orchestration and agentic workflows
- tool-using LLM systems
- fast prototypes and production-ready MVPs
- Python, TypeScript, Next.js, Convex, SQL, AI APIs
- Agent Orchestration Protocol — open-source framework for multi-agent AI workflows and CLI tooling
Data science and technical roots. More recently: real-time anomaly detection, LLM-based automation, and end-to-end AI product development.
Applied AI engineering, product collaborations, and freelance work across Europe or remote.