OVERVIEW
Finding the right product is not always easy when the starting point is just a photo. To solve that, the client explored how AI could be used to recognize products from images and match them against a furniture catalog.
The original goal was to help customers identify products or similar alternatives more easily and reduce friction in the buying journey. As the work evolved, the scope expanded into an internal use case as well — where employees working with circularity could use the same capability to identify returned products faster.
This turned the solution into more than a customer-facing feature. It became an example of how one AI capability could support both customer experience and operational efficiency across multiple workflows.
A strong validation of the work came when the solution was presented at Google I/O in San Francisco — giving external recognition to both the technical quality and the innovation behind the product.
OUR APPROACH
We started with a clear user problem, built something usable, and expanded the value as the capability proved relevant across multiple workflows.
Computer Vision
We built an application powered by an AI model trained to recognize products in the client's catalog based on images. A user could take a photo and get back the closest matching product.
Customer Experience
The product was initially designed to support customers trying to identify furniture or find similar items more easily — reducing friction in the buying journey.
Circularity Use Case
The same AI capability was extended to support internal teams working with circularity, helping them identify returned products with less manual effort.
Practical AI
This case shows how we approach AI in practice — start with a clear user problem, build something usable, and expand the value when the capability proves relevant across workflows.
OUR RESULTS
One AI capability that created value across both customer experience and internal operations.
Product Identification
The result was a working AI-based application for image-driven product identification — practical, fast, and integrated into real business workflows.
Dual Value
The solution demonstrated how AI could simplify product discovery for customers while also supporting internal teams handling returned goods through the same core capability.
Google I/O Recognition
The solution was presented at Google I/O in San Francisco — giving external recognition to the work and reinforcing its strength as an innovation case.
Flexible AI
By extending the same core capability into multiple workflows, the project showed the value of building AI features that are practical, flexible, and tied to real business needs.