About

AI builder working at the intersection of agentic engineering and real-world urgent-care operations — learning in public.

I build AI systems and write about what I learn doing it. My work sits at the intersection of two worlds that rarely meet: agentic engineering — LLM-driven agents, RAG, and the infrastructure that makes them dependable — and urgent-care operations, where those systems have to hold up against messy, high-stakes, real-world demands.

This site is my notebook in the open. The blog is for essays and field notes — things I figured out the hard way and want to remember. The garden is rougher: living notes I tend over time, from early seedlings to evergreen ideas. The now page is what has my attention this week.

What I write about

  • Agentic engineering — building reliable AI agents, RAG pipelines, and the scaffolding around them.
  • Operations under load — running real workflows where AI meets the constraints of a working business.
  • Building in public — the tools, infrastructure, and habits that compound over time.

Everything here is generalized — principles and lessons, not anything specific to a client or patient. If something resonates, find me on GitHub or LinkedIn.

Live board

The stack I build with

14 tools
  • TypeScript Languages default for everything
  • Python Languages ML + data work
  • Astro Web this site
  • React Web
  • Cloudflare Infra Pages + DNS + Workers
  • Postgres / pgvector Data retrieval store
  • Neo4j Data knowledge graphs
  • LightRAG AI graph RAG
  • Claude AI agentic workhorse
  • Whisper AI speech-to-text
  • tmux Tooling fleet of agents
  • Obsidian Tooling second brain
  • GitHub Actions Infra CI/CD
  • Docker Infra