Six shipped apps on Farcaster — a trust-based freelance marketplace, a collaborative art relay, a 3D runner, a community wiki, a trading journal. All solo, all open-source, all self-taught.
Every project started with a specific person's problem — a friend needing work, a community wanting a game, a habit worth tracking. Built end-to-end: frontend, backend, and deployment.
A freelance marketplace built on Farcaster's social graph — so people can hire and get hired by others they already trust, instead of strangers. Built after a friend needed work done and had no safe way to find someone. Next.js and OnchainKit handled the Farcaster mini-app integration; Redis powered real-time job and application notifications.
A daily collaborative art relay: one person posts an image and passes the turn; each next person gives an AI image editor a single prompt before passing it on. After 10 edits, the chain becomes an auctionable piece. Postgres tracked each chain's state and turn order; Gemini's API handled the actual image edits.
A 3D endless-runner mini-game built for the $BETR community — play as a cat collecting letters, with a live leaderboard and daily-attempt limits to keep it fair and competitive. Three.js rendered the 3D runner in-browser; Neon Postgres kept the leaderboard live and race-condition-safe under concurrent plays.
A community-editable wiki for Farcaster projects and tokens — think Wikipedia, scoped to the ecosystem, with contributor history and reputation scoring built in. Supabase handled auth and versioned content storage; the Neynar API pulled live Farcaster profile and social data.
A minimal trading journal that makes you commit before you trade — entry, stop-loss, and take-profit levels locked in, with risk/reward math done for you and a history of wins and losses over time. Firebase stored trade history and synced across sessions; Wagmi let traders mint their plan on-chain as a public commitment device.
An experimental text/image-to-video mini app for Farcaster — in progress, and the current edge of what I'm learning.
I scope, architect, and ship full products working closely with AI coding tools — from data model to deployment. I can read and modify most of this code myself, and I'm actively building my independent debugging skills project by project. The technologies below are what I've shipped real, working products in.
Open to internships and early opportunities in software — happy to talk about any of the projects above, or start a new one.