Fashion tech glossary
Short definitions for the ideas behind Velvet and modern fashion software—virtual try-on, agentic workflows, and more.
Personal fashion OS
A personal fashion OS connects discovery, body-accurate preview, and purchase decisions in one loop—idea to cart with less guesswork. Velvet is building that layer for fashion rather than a single disconnected feature.
Digital Double
A Digital Double is a body-accurate model for previewing clothes—not a generic avatar. The goal is a preview aligned to your proportions so what you see matches what shows up.
Body twin
Informal term for a personalized body model used in try-on or fit tools. Velvet uses Digital Double to stress accuracy and shopping use cases over gaming-style avatars.
Virtual try-on
Uses computer vision and graphics to show how a garment might look on you—often from a photo—before purchase. Quality hinges on faithful body shape and fabric drape, not only a fast render.
Learn moreAgentic commerce
Shopping software that takes goal-directed actions for you—searching, comparing, curating—inside limits you set, instead of only returning static search results.
Learn moreConversational commerce
Product discovery through chat: mood, occasion, budget, and taste in natural language rather than fixed keyword search and filter grids.
Learn moreAI personal stylist
Suggests outfits or pieces from taste, context, and constraints. It works best with real catalog data and a way to see items on you—such as a Digital Double.
Fashion AI
Spans recommendations, visual search, styling dialogue, and fit-related models. It pays off when grounded in real inventory, clear sourcing, and honest return policies.
Fit prediction
Estimates how a size or cut may work from measurements, past buys, or returns. It pairs with virtual try-on for confidence but does not replace seeing proportion on your body.
Return rate optimization
Improving sizing guidance, imagery, and expectations to cut unnecessary shipments and reverse logistics—where accurate try-on and fit signals matter.