Amazon licenses AI shopping technology to retailers, signs Kate Spade parent Tapestry as early adopter

Amazon is doing what Amazon does best: selling the picks and shovels. The company is now licensing its AI-powered shopping technology to other retailers through AWS, turning its internal retail expertise into a cloud product that competitors can buy off the shelf.

One of the earliest and most prominent adopters is Tapestry, the parent company behind Kate Spade New York, Coach, and Stuart Weitzman. Tapestry has built a generative AI engine on top of AWS using nearly 20 different services, and the results so far suggest this isn’t just another enterprise software pilot that quietly dies after the press release.

What Tapestry actually built

The centerpiece of Tapestry’s AI deployment is an application called “Tell Rexy,” a tool designed for frontline store associates rather than back-office executives. In its first year of operation, Tell Rexy collected nearly 30,000 pieces of feedback from store employees across most North American Coach locations.

The app runs on Amazon Bedrock, AWS’s managed service for building generative AI applications. It also uses Amazon Transcribe for speech recognition and Amazon Translate for multilingual support, which matters when your store associates speak different languages across various markets.

Tapestry built the entire solution in just eight weeks, working with AWS Professional Services. The company claims the reusable components they developed accelerate generative AI application development by a factor of 10.

Tell Rexy is currently active in most North American Coach stores and is expanding to Kate Spade New York locations.

Tapestry has also secured its first AI patent for a platform called Mira, described as an AI data analytics tool.

The broader AWS retail play

AWS’s Just Walk Out technology, the checkout-free system originally developed for Amazon Go stores, has been implemented in over 180 third-party locations worldwide as of September 2024. That rollout established a template: develop the technology internally, prove it works, then package it for external customers.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.