I've spent 16 years building engineering organizations and shipping software in healthcare, retail, and cloud infrastructure. I tend to work in industries where the software actually matters to the people using it.
I lead engineering at Walmart Health & Benefits Technology. I own two portfolios across a platform that serves 1.6 million members and runs about $6 billion in spend. One is a builder platform that lets business and engineering teams ship without waiting on central teams. The other is a new healthcare business line I built from scratch — new domain, new team, no existing code.
At AWS, I built and ran a 50-person engineering organization. We owned the contract management and analytics systems behind roughly 70% of AWS revenue. The work was internal-facing but high-stakes — if it broke, deals stopped moving.
At MTM Health, I was the first data science and analytics hire. I built the AI and data systems that helped grow the business from $600 million to $900 million in revenue. When I started, there was no data scince engineering team. When I left, there was.
A prototype that uses AI agents to navigate healthcare claims data on behalf of members. Instead of scripted chatbot flows, agents reason over 835 EDI transactions, denial codes, and member records to actually explain what happened with a claim and what to do next. Built with Python, LangChain, Streamlit, and React.
I ride bikes. This pulls my ride data from Strava, recovery data from Oura, and weather from Open-Meteo into one place. I ask questions in plain English and get answers grounded in real power and recovery numbers — not generic advice. TypeScript, Supabase, Telegram as the interface. Still private while I get the core right.
An AI assistant for short-term rental companies. Early stage — more details coming soon.
GitHub's traffic API only keeps 14 days of data. This archives it. Zero dependencies, daily cron, JSON archive, static dashboard. Small tool, done properly.
Languages & Frameworks: Python · TypeScript · React · Node.js · Streamlit
AI/LLM: LangChain · RAG · Vector databases (Pinecone, pgvector) · MCP · Prompt engineering · Claude API · OpenAI API
Data & Integration: PostgreSQL · Supabase · Kafka · REST APIs · Webhooks · ETL / data pipelines
Cloud & DevOps: AWS · Docker · Terraform · CI/CD · GitHub Actions
Domain: Healthcare EDI (835/837) · HIPAA · Pharmacy Benefits (NCPDP D.0) · Claims adjudication · Property management / STR operations
Healthcare is fragmented. Most of the complexity patients deal with exists because systems don't talk to each other. I think the teams that fix that — with good software, not more process — will define the next decade of the industry.
I still build. Leaders who stop building eventually stop understanding what builders need.