📍 Los Angeles, CA | 💻 Full-stack software engineer with a strong product sense | 🚀 0→1 builder at startups

Currently at an AI startup, where I've helped scale the business to $1M ARR. Outside of work, I'm going deep on agentic engineering — building full apps end-to-end with AI agents.

• 10 years of experience as a full-stack software engineer
• Lead maintainer of react-native-elements, the cross-platform React Native UI toolkit (25.8k+ stars) —
• Architected cross-platform subscription infrastructure processing $100K monthly revenue
• Built and scaled notification infrastructure processing 20K+ daily notifications, improving open rates by 14% through personalization
• Created and taught MobileSpace, a free, in-person course at SF State that taught students mobile development (iOS, then React Native) to fill a gap in the CS curriculum and help them land internships and SWE roles (2016–2018)
• Built @nftsalesbot (nft-sales-bot), a Twitter bot posting single-tweet stories of NFT flips and fumbles on the OpenSea API — grew it to 25k+ followers
• 1.5k+ GitHub stars across personal projects

The net-new work. Each project is a different shape of app, built end-to-end with AI agents — the goal is learning the tools and scaffolding that agentic engineering takes: skills, hooks, sub-agents, eval harnesses, and CI that agents can drive.

• 🔁 stuck-problems-agent - An autonomous research loop for important-but-stuck problems, run by a scheduled Claude Code routine. Scout, planner, generator, evaluator, and publisher sub-agents ��� the repo itself is the policy, the assets, and the memory.
• 🏛️ history-stories - A 3D museum you can walk through in your browser, where every painting is a short story from history. Comes with its own content-agent harness — planner and evaluator sub-agents, skills, and evals. (historystories.vercel.app)
• ⚔️ claw-debates - AI vs AI debate arena: watch two LLMs argue opposing sides of any topic with a real-time reveal flow, an automated judge, and audience voting. Built to explore how models reason under adversarial conditions. (clawdebates.vercel.app)
• 🧠 quiz-me - Multi-user trivia log where Claude quizzes you on topics you pick and grades honestly. Next.js + Neon Postgres + Anthropic SDK, with CI-driven database migrations. (quizmenexus.vercel.app)
• 📚 deep-research-agent - Long-form research essays co-written with an AI agent through a documented brainstorm → research → draft → publish workflow, with open data, charts, and annotated source bibliographies.
• 📊 cursor-stats - Upload your Cursor AI usage CSV and ask questions in natural language to get interactive charts. Zero-hallucination architecture — the LLM picks the query, the client resolves the data, and your CSV never leaves the browser. (cursorstats.vercel.app)
• ⛳ math-golf-game - Puzzle game that uses math operations as golf clubs — multiply, divide, add, and subtract your way to the target in as few strokes as possible. (mathgolfgame.vercel.app)

More agentic experiments are in the works — I'll open-source them and add them here as they're ready.

The proven work — React Native components used in production by other people's apps.

• 🧰 react-native-elements (25.8k+ stars) - Cross-platform React Native UI toolkit — lead maintainer and longtime contributor
• ⭐ react-native-ratings (940+ stars) - Tap and swipe ratings component for React Native
• 🌄 react-native-parallax-scrollview (270+ stars) - The Parallax ScrollView component we all deserve