Industry
Furniture retail
Shoppers cannot picture the room, so the sale waits or returns
The problem
Furniture is bought on imagination. A shopper cannot tell whether a sofa suits their living room until it is in their living room, and that uncertainty is where the sale is lost — or comes back later as a return.
The usual answers do not close the gap. A room set in a showroom is not the customer's room, a dimension in centimetres is not a picture, and an augmented-reality viewer asks the shopper to do the work of imagining anyway, on a phone, in the room, while holding it up.
What we built
The AI Space Visualizer: a shopper photographs their room and sees the retailer's real, in-stock products placed in it, revises the result by describing the change, and sends the finished design to the retailer's sales team with the products already identified.
Read the AI Space Visualizer case studyWhat transfers
The catalogue is the constraint, not the prompt
A generative model will produce a beautiful sofa the retailer does not stock, and that output costs more than a broken one — it produces a shopper who has decided and a sales team who cannot fulfil. Placement has to be decided against real inventory before generation is asked for anything.
The hand-off is the product
A visualiser that ends at a picture has entertained somebody. One that ends with a named product list in front of a salesperson has moved a sale, and the difference is entirely in what happens after the image renders.
Retailers rarely have catalogue data ready
Photography is inconsistent, dimensions are missing, colourways are recorded loosely or not at all. Any system that assumes a clean catalogue will stall on contact with a real one, so the tolerances belong in the design rather than in an onboarding document.
Capability: Generative & Visual AI