Per supplier
Storefronts
platform · 2025 · client unnamed
Multi-vendor marketplace where every supplier gets a storefront and the catalogue carries the material-specific attributes tiles and furniture actually need.
Per supplier
Storefronts
2
Surfaces
API plus admin dashboard
The problem
Ceramic tiles and furniture are not generic products. Tiles have finish, wear rating, dimensions and batch; furniture has material, dimensions and assembly. A generic product schema forces both into attributes that match neither, and buyers cannot filter on what they actually care about.
The approach
Modelled category-specific attributes properly instead of flattening everything into a shared product shape, and gave each supplier their own storefront within the marketplace. Added AR asset delivery so a buyer can place a piece in their own room, plus AI-assisted catalogue search.
Architecture
Category-aware product attributes rather than a single flattened schema, with per-supplier storefronts and separate asset handling for AR models.
Each with what it cost
A decision without a stated tradeoff is a preference.
01
Why
Buyers of tiles filter on things buyers of sofas have never heard of. One shared attribute set means both categories are described badly.
What it costs
Category-aware validation and filtering is more code than one generic product model.
Outcome
Buyers filter on the attributes that matter for the material, suppliers manage their own storefronts, and AR turns a catalogue photo into something a customer can evaluate in their own space.