Amazon Exports Its AI Shopping Engine to Third-Party Retailers

Retailers can now deploy conversational shopping agents built on Amazon’s proprietary infrastructure, with Kate Spade leading the rollout. By leveraging the same architecture that generated nearly $12 billion in incremental sales for Amazon last year, brands aim to replace static keyword searches with personalized, intent-driven gift concierges.

The service provides companies with the necessary architecture, starter code, and AWS integration support to launch AI assistants in weeks rather than years. Built upon Amazon Bedrock, AgentCore, and OpenSearch, these tools allow retailers to fine-tune the shopping environment to match their specific brand voice and customer base. Amazon reports that conversational sessions currently yield conversion rates 3.5 times higher than traditional search methods.

Kate Spade, a subsidiary of Tapestry, serves as the primary use case for this technology. Their new AI Gift Concierge addresses the friction of holiday and occasion-based shopping, a category where 53% of consumers report significant stress. Fabio Luzzi, chief data and analytics officer at Tapestry, noted that the tool was developed specifically to resolve identified consumer pain points during the discovery phase. Following a ten-week testing period, the assistant now guides shoppers through product catalogs based on occasion and individual input.