I'm a software engineer who works across the entire life-cycle of a product, from interfaces to backend services to AI systems, currently building agentic architectures and LLM-powered workflows in a production enterprise environment.
I'm naturally curious and that curiosity spills into everything I do. When I work on a product, I genuinely want to understand it: what problem it solves, who uses it, and why it matters. This makes me ask a lot of questions and I think that it helps me to be a better engineer.
Always building something on the side. Right now:
-
DocForge · Live · Platform · Layout library An agent-friendly document format and the platform around it, built and deployed end to end: an open-source TypeScript layout library, a visual editor, and an MCP server with 40+ tools where people and AI agents create, review, and publish documents together.
-
CopaSim · Play A persistent multiplayer football management game with a deterministic simulation running a world of 662 clubs and 23,279 players in the host's browser, fed by a 26-stage Python data pipeline.
-
Webrolê · Play An online party-game platform I designed and built solo: 13 minigames, rooms of up to 20 players, peer-to-peer voice and video over WebRTC, real-time sync over WebSockets, and a Discord Activity integration.
-
temporal-adk-agents An architecture for building agents on Temporal.io with the new ADK integration.
- PostHog: AI Observability
- Fix span highlighting when deep-linking LLM traces: sidebar compared the URL's span ID against the internal event UUID; added a selector that resolves the canonical ID, with unit tests.
- Restore notable properties table on OpenAI install docs: the page announced a properties table that never rendered; wired up the missing snippet to match the other provider pages.

