Yes. Elara's occasion reasoning distinguishes between wedding sub-events (mehendi, sangeet, reception, puja), festivals (Diwali, Eid, Navratri, Holi), and everyday occasions. The formality and aesthetic implications of each are built into Elara's reasoning — a sangeet outfit is different from a reception outfit, even for the same shopper. Catalog metadata enrichment (adding occasion tags to products) makes these distinctions sharper.

Yes. Elara handles the full spectrum from traditional (pure silk sarees, heavy embroidered kurtas) to contemporary fusion (Indo-western coordinates, minimalist ethnic silhouettes, modern bandhani). The Style Graph learns which aesthetic register each shopper prefers, so repeat visitors get recommendations aligned with their taste rather than the store's full range.

VTO works best for structured silhouettes — anarkalis, salwar kameez, coordinates. Sarees and heavily draped garments have lower accuracy due to the complexity of draping physics. Elara flags these automatically. For drape-heavy garments, VTO shows the garment on a model alongside the shopper's try-on attempt, giving context for what the draping would look like.

Yes. Elara works across your full catalog. When a shopper specifies an ethnic or fusion occasion, Elara draws from relevant pieces across your catalog. When a shopper's brief is western, it responds accordingly. The Style Graph tracks which aesthetic register each shopper gravitates toward, so the recommendations over time become increasingly accurate to each shopper's preference pattern.

Yes. The SDK works with Hydrogen and any front end that can load a script or call the API.

Your catalog, stock and anonymous on-site behaviour. No personal data is required to style a shopper; nothing is shared across brands or used to train models for anyone else.