I ship production systems end to end — and I build the tooling that lets me do it fast. My work sits where careful backend engineering meets AI automation: distributed services that have to be correct, and agent-driven workflows that make shipping them quick.

• A large multi-tenant B2B SaaS platform — payments and ledger systems, third-party data integrations, permissions, and compliance reporting across a Python + TypeScript microservice backend
• AI automation pipelines — LLM-driven enrichment, personalization, and response routing wired into CRM and email infrastructure
• Event-driven systems — a multi-tenant email automation engine with a configurable rule engine, templating, and a full audit trail behind every decision
• Developer tooling — Playwright screenshot harnesses and agentic dev workflows: worktree-per-change, automated verification, AI-assisted review

• Verification before completion — a change isn't done until I've run it, exercised it, and captured the evidence
• Agentic by default — I lean hard on AI coding agents, with custom skills, hooks, and harnesses built around them
• Isolated workspaces — one worktree per change, so parallel work never collides

Languages · TypeScript · Python · JavaScript · Ruby · SQL

Frontend · React · Next.js · Vue · React Native · Tailwind

Backend · Node.js · Express · FastAPI · REST · WebSockets · microservices

AI · Claude API · OpenAI · Gemini · agentic workflows · RAG / vector search · n8n

Data · PostgreSQL · Supabase · Firebase · vector databases

Infra · Docker · AWS · Playwright · GitHub Actions

📍 Los Angeles, CA