Mannequin
RUNNINGYour wardrobe, photographed once — dressed by AI every morning.
70%
Expo
React Native
TypeScript
Supabase
Claude
Replicate
Mannequin turns your real closet into a daily outfit assistant. Photograph your clothes; AI removes the backgrounds and auto-tags each garment (category, colors, pattern, warmth, occasions). Every morning the app suggests three weather-aware outfits assembled from your own wardrobe — and learns your taste from likes and wear history.
How it works
- An npm-workspace monorepo: Expo Router app, feature modules, and
@mannequin/engine— a pure, dependency-free TypeScript outfit engine. - The engine is fully deterministic (seeded by user, date, and shuffle index), so the phone and the server-side notification cron compute byte-identical suggestions: the morning push shows exactly the outfits the app will.
- Supabase handles auth, Postgres with RLS, storage, and ten Deno Edge Functions; garment tagging runs on Claude with background removal via Replicate, measured by an eval harness against labeled photos.
Status
Pre-launch and in active build, targeting an Indonesia-first beta. Freemium: free up to 40 garments, Premium unlocks unlimited closet, tomorrow preview, week planner, and AI visualization.