Sr. Technical Product Manager based in Seattle. I spent 10+ years as a software engineer before moving into product and I never really stopped building.
At Amazon and Moz, I've led AI/ML products, data platforms, and developer-facing tools from zero to launch — shipping sentiment platforms across 245K+ feedback touchpoints, growing products to $73K ARR in beta, and cutting prototype cycles from months to weeks. I'm most effective working at the intersection of technical depth and customer outcomes: going deep enough to earn engineering trust while keeping the work anchored to what actually matters.
Outside of work I build personal AI projects, write about what I'm learning, and play D&D with my kids.
A household AI agent connecting Monarch Money to a shared Telegram interface, running persistently on a home Mac Mini with zero cloud costs. Dual-mode: scheduled reports push proactively, a tool-driven layer handles ad hoc questions. Built on MCP, LangChain, and APScheduler.
LLM-powered D&D Dungeon Master for family tabletop sessions — voice synthesis, structured session state, enforced dice randomness. Multi-modal design follows players from desktop to terminal to mobile. Shipped to real users. Written about on the blog.
Food-first AI pairing tool deployed on AWS Lambda via SAM. Multi-provider LLM support, response caching. Documented in a four-part build series.