01
Visual Search
Upload any image and find visually similar items using FashionCLIP embeddings trained on fashion data.
02
Text Intelligence
Natural language queries with NLU parsing. Extracts color, gender, occasion, and price range automatically.
03
Hybrid Ranking
CLIP semantic scores fused with BM25 lexical recall for the best precision and coverage across all queries.
04
Outfit Discovery
Complementary item suggestions to find pieces that complete your look, powered by semantic similarity.
The Search Engine
Find it with
words, images,
or both.
Powered by FashionCLIP, FAISS, and BM25. Multimodal fashion retrieval at scale.
Upload