Getting Started
How VisualDNA™ Works
Search, Collections & Merchandising
Dealbreaker™ Filters
Bundles & Basket Growth
Shop Similar
Brand, Content & Experience
Recovery & Retention
Pricing & Billing
IntentWeave™: AI product attributes without manual tagging
IntentWeave™ reads product imagery and converts visible characteristics into structured catalogue signals such as silhouette, neckline, sleeve, texture, pattern and colour family, reducing the need for manual tagging and helping new products enter discovery faster.
Why product attributes matter
Visual discovery needs more than a broad product category. Shoppers often make decisions around characteristics such as neckline, sleeve, silhouette, colour, texture and fit cues.
What IntentWeave™ does
IntentWeave™ analyses product imagery and generates structured signals that RecoMelon can use across recommendations, filters and search.
Examples of attributes
- Silhouette and shape.
- Neckline and sleeve characteristics.
- Colour family and pattern.
- Texture and visible material cues.
- Category-specific visual characteristics.
Does this eliminate all merchant product data?
No. Merchant data remains useful for commercial facts and business rules. IntentWeave™ reduces dependency on manually creating every visual merchandising attribute before discovery can work.
How IntentWeave™ supports Dealbreaker™ Filters
Structured visual attributes give RecoMelon a language for shopper refinement. A shopper can preserve the product characteristics they like while changing the characteristic that makes the product unsuitable.
How it supports The Index™
Attribute intelligence can help The Index™ interpret natural-language requests that describe product characteristics rather than exact product names.
