Background

I build AI developer tools. MCP servers, Claude Code skills, agentic workflow frameworks, and collaboration safety engines. I design how AI coding agents should coordinate, share context, enforce quality gates, and hand off work. I manage full product development — concept, design, architecture, build, and ship — using AI coding agents as the engineering team. Currently shipping these tools while serving as course architect for Red Hat's flagship enterprise automation training.

Before AI tooling, I built products end-to-end. Seven years as a product manager, from defining technical vision and architectural roadmaps to taking an entire networking product line from concept through hardware specs, cloud platform, and customer-facing UI. Grew one product line's revenue 4x. But the gap between design and implementation is where good ideas get diluted — so I started closing it myself.

I've been building things since before the web was a business. Co-founded my first company at 18. Enterprise architecture, one of the first fully virtualized U.S. data centers, a cloud delivery platform before "cloud" was a word, and engineering management across networking, security, and infrastructure teams. The thread through all of it: spot the problem, design the solution, ship it.

AI as a force multiplier. I manage agentic coding tools as a development team — writing specs they implement, reviewing their output, enforcing standards. I created Project Foundry, an open-source methodology for building production-grade software from day one with AI agents. I've built an MCP server for multi-agent coordination, a control plane that orchestrates Claude Code and Codex, and a safety engine that enforces the Four-Eyes Principle. Every concept is mine. Every design is mine. The git histories prove it.