AI search: shoppers describe what they need or take a photo of the product

Shoppers don't always search by product name. They type what they need — "something for leg cramps" — or they hold the empty package in their hand. From today Findora understands both.

Semantic search and photo search, in the same search bar.

Semantic search

When a shopper types a sentence and presses Enter, or the regular search finds few or irrelevant results, AI "translates" the question into product terms from your catalogue and keeps only the products that fit:

• "something for cramps" → magnesium;
• "chapped lips" → lip balm;
• "hair dye under €15" → hair dye, with the price limit.

The results show "Understood: … (AI)". Only products from your store appear; if you have nothing relevant, unrelated products are not shown. Your "categories not recommended" (e.g. medicines) stay out.

Findora's regular search already understands many such questions on its own; AI steps in where it falls short.

Search by photo

A camera button appears in the search field. The shopper takes a photo of a product (or picks one) and AI reads the package — brand, name, size, barcode if visible — and finds the same product in your catalogue; below it, under "Similar products", items of the same type from other brands (e.g. other lip balms), not everything else the same brand makes. The results show "From the photo: …".

The photo is resized on the shopper's device before it is sent and is not stored anywhere.

In detail: semantic search and search by photo.

Plans and billing

• Available in the Pro and Enterprise plans.
• Charged to the store's AI credits: semantic search only for each new question (the same question within 30 days is free), photo search per use.
• Turn them on or off in the portal: Settings → Features. If AI credits run out, they are hidden and search keeps working normally.

Want to see how it would work with your own catalogue? Request a demo or get started now.