From customer signals to better fashion choices.

Two connected case studies: a full-population recommendation system and a conversational product-discovery experience built on the same fashion catalogue.

H&M Personalized Fashion Recommendations

Complete

A 17-signal retrieval and ranking system that creates twelve personalised fashion recommendations for 1.37M customers.

Recommender systemsRetrieval + rankingGCP

AI Fashion Search & Shopping Chat

Complete

Semantic product discovery and LLM-assisted shopping chat, built on the same H&M fashion catalogue.

Semantic searchLLMPersonalisation

Try it yourself

Search the catalogue. Then talk through the choice.

A public, catalogue-grounded demonstrator makes the search and shopping-chat layer tangible—without exposing customer data.

Open the live experience