Smart wardrobe assistant that connects to a user's calendars, location, and wardrobe inventory to recommend outfits tailored to events and weather.
For the first milestone we are focusing on:
- Quick-start onboarding: new users upload short closet videos or a small set of photos and see their digital wardrobe appear over time.
- Guided item capture: optional step-by-step flow for high-accuracy item capture.
- Manual cleanup: simple tools to edit, merge, or delete items after automated import.
- Backend: Python + FastAPI, PostgreSQL for persistence.
- Vision worker: Python pipeline (YOLOv8 + EfficientSAM + CLIP) running as a separate service for media ingestion and wardrobe extraction.
- Client(s): initially a simple web client for onboarding and wardrobe review (mobile-friendly), with native mobile capture planned later.
See docs/onboarding.md for detailed onboarding flows and docs/onboarding_wireframes.md for low-fidelity wireframes.