What challenges did Tapestry face before AI implementation?

Store feedback collection was fragmented and could not be scaled across a large retail network. Corporate teams only had anecdotal store-visit information and an incomplete view of customer trends and associate needs. Manual collection, summarization, and analysis of associate feedback was not practical at Tapestry's scale.

What was the impact of the AI solution?

Tell Rexy is live across most Coach stores in the United States. Several thousand associates have used the application and provided nearly 30,000 pieces of feedback in one year. Tapestry says it can spin up new AI applications 10 times faster thanks to the reusable architecture. The company is now expanding the applications to Kate Spade and seeing interest from other business units.

Source

Scaled productionEvidence: Medium50/100

Tapestry builds generative AI feedback and analytics chatbot on Amazon Bedrock

Use case typeVoice automationUpdated Jan 31, 2025

Tapestry, the luxury fashion holding company behind Coach, Kate Spade New York, and Stuart Weitzman, built an in-house generative AI engine on AWS. The solution helps collect, synthesize, and analyze store associate feedback at scale across a large retail network. Two applications, Tell Rexy and Ask Rexy, support voice feedback collection and analytics question answering for store and corporate users.

Organization: Tapestry
Industry: Retail
Location: United States
Published: January 2025

Reported outcomes

10x

application development speedupTime & speed

30,000 countfeedback entries collected

Strategic outcomes

Cost efficiencyScaled associate feedback collection across most Coach storesBetter decisions & insightImproved visibility into store operations and customer preferencesMarket & geographic expansionPreparing to extend the solution to additional brands and business unitsOther strategic outcomeCreated a reusable foundation for future AI applications

Catalog median for time & speed deployments: +60% across 143 reported metrics. Compare benchmarks →

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Feedback entries collected: 30,000 count increase

AWS BlogJan 31, 2025Blog postExplicit claimMedium evidence strength

“provided nearly 30,000 pieces of feedback in 1 year”

Normalized claim

Application development speedup: 10 x increase

AWS BlogJan 31, 2025Blog postExplicit claimMedium evidence strength

“the company reports that it can spin up new applications 10 times faster”

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Tapestry, Coach, Kate Spade New York, Stuart Weitzman
Provider: AWS
Maturity: Scaled Production
Linked source: AWS Blog

The solution helps collect, synthesize, and analyze store associate feedback at scale across a large retail network

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Voice automation
2Conversational analytics
3Workflow orchestration

Store feedback collection was fragmented and could not be scaled across a large retail network.
Corporate teams only had anecdotal store-visit information and an incomplete view of customer trends and associate needs.
Manual collection, summarization, and analysis of associate feedback was not practical at Tapestry's scale.

Tapestry built a generative AI engine on nearly 20 AWS services with Amazon Bedrock at the core.
Tell Rexy runs on store devices and uses Amazon Transcribe for speech-to-text, Amazon Translate for multilingual input, Amazon S3 for storage, Amazon Comprehend for sentiment analysis, and Amazon Athena for queryable analytics tables.
Ask Rexy uses retrieval-augmented generation and text-to-SQL to answer questions from corporate analysts, combining semantic retrieval over content in Amazon S3 with SQL generation for Amazon Athena.
The company designed the engine with reusable components and extensible architecture so it can support additional AI applications across the business.

Technologies

Amazon Bedrock Amazon S3 Amazon Transcribe Amazon Translate Amazon Comprehend Amazon Athena

Tell Rexy is live across most Coach stores in the United States.
Several thousand associates have used the application and provided nearly 30,000 pieces of feedback in one year.
Tapestry says it can spin up new AI applications 10 times faster thanks to the reusable architecture.
The company is now expanding the applications to Kate Spade and seeing interest from other business units.

Architecture

Tell Rexy is deployed on store devices and converts spoken feedback to text with Amazon Transcribe, translates multilingual inputs with Amazon Translate, stores and processes feedback in Amazon S3, and updates Amazon Athena tables for analysis. Ask Rexy uses Amazon Bedrock-driven retrieval-augmented generation and text-to-SQL over Amazon S3 and Amazon Athena to answer analyst questions. The broader engine is built as a reusable, extensible foundation across nearly 20 AWS services.

Sources & evidence1

Evidence: Medium50/100Evidence strength