Raincoat

Product Demo

Smart outfits.
Zero privacy compromise.

Upload your closet once and get daily, weather-based outfit suggestions. All image processing runs locally. Your photos never leave your device.

Your photos stay private

Raincoat runs background removal, detection, and embedding generation on-device with ONNX Runtime Web and WebGPU. We only receive anonymized vectors and user-validated tags.

In development • Client-side ONNX pipeline

Everything you need.
All on-device.

Smart features that respect your privacy

Smart Auto-Tagging

Automatic clothing classification and attribute detection to eliminate manual tagging and speed up closet organization.

Weather-Based Suggestions

Personalized outfit recommendations matched to local weather and forecast data so you're always prepared.

Similarity Search

Find visually similar items across your wardrobe and discover alternatives from partner catalogs using vector search.

Multi-Location Support

Manage multiple closets for different cities, offices, or travel destinations with location-aware recommendations.

Instant Client Processing

ONNX models running on device with progressive loading and WebGPU acceleration for near-instant results.

Travel Mode

Pack smarter with suggested packing lists and outfits tailored to the forecast at your destination.

Built with modern tech

Privacy-first architecture powered by on-device inference and efficient vector search.

Next.js
Ruby on Rails
PostgreSQL + pgvector
ONNX Runtime Web
YOLOv8
FashionCLIP
U²-Net
WebGPU
TypeScript