Tapestry Is Scaling AI Across Commerce, Design, Operations
Tapestry, which manufactures the Coach and Kate Spade brands, is making its products purchasable directly from Google search, including AI mode and the Gemini app.
Consumers browsing a relevant product while using one of these engines will now see a "Buy" button appear within the search interface, removing the added step of having to navigate to the DTC site and streamlining the path to purchase.
The company is using a Universal Commerce Protocol, which is an open industry standard for agentic commerce that is supported by several tech and e-commerce companies. Each transaction will be secured through Google Pay.
The capability is currently live in the U.S., and the company plans to expand to other regions in the future.
Tapestry CEO Joanne Crevoisera said consumer behaviors continue to evolve, pointing to when discovery previously happened in a store window or a magazine before it switched over to social media.
"Now it's happening inside a conversation with AI," he said in a statement. "Our goal is for shoppers to find and engage with our brands at the moment when inspiration strikes, whether it be when a consumer passes by one of our stores, peruses our websites or chats with AI.”
Chief Information and Digital Officer Yang Lu said Tapestry built this capability to function within its existing operations, but emphasized that it is an agnostic approach that will allow the company to move into new AI environments as consumer behaviors shift.
"That's possible because of the strategic investments we've made in technology capabilities and infrastructure as well as the strength of our data and consumer insights engine," he added.
How AI Is Touching Multiple Areas of Business
The launch is part of Tapestry's broader AI strategy, which focuses on driving company growth and enabling transformations across teams and operations.
The company is on an accelerated journey to execute AI-enabled business optimizations that address both back-end operations and front-end consumer experiences.
At the core of its enterprise strategy is the newly patented Mira platform, built alongside the company's global data fabric. It is designed to consolidate operational data into a single ecosystem, streamlining decision-making across inventory management and assortment planning.
By replacing manual legwork and fragmented dashboards, it compresses analysis times from several days down to seconds or minutes, giving teams the agility to rapidly react to shifting consumer trends while maintaining strict role-based access controls.
The company has also deployed customer-facing AI innovations, including the Kate Spade AI Gift Concierge. Built on AWS's Bedrock AgentCore platform, the conversational agentic commerce tool provides personalized product recommendations by analyzing gift recipient traits, event types and shopper intent. The tool also taps into historical shopping behaviors from Amazon’s Alexa for Shopping platform while preserving Kate Spade's distinct brand voice.
And more recently, the company partnered with Raspberry AI to connect the fashion product lifecycle in a single continuous workflow. The platform, which is now enabled by agentic capabilities, will allow Tapestry to more quickly move from consumer and market intelligence and trend research through design, technical design, merchandising, wholesale, marketing and e-commerce processes. The company's designers, merchandisers and marketers can provide input, providing consumer insights, moodboards and fabrics from their libraries to develop and iterate production-ready technical designs and visualize complete assortments.
Sharing Teachable Moments
These initiatives reflect Tapestry's progress in building a scalable blueprint for AI integration by pairing institutional knowledge with agentic technology. And the company has been vocal in sharing both its accomplishments and roadblocks along the way.
Last year at NRF’s Big Show, Karthik Ramakrishan, director of planning systems and RPA at Tapestry, said the biggest challenge is that agentic AI models need a space in which brands can teach — not necessarily train — in a low-effort, non-time-intensive way. Otherwise, prompts return incorrect information because the model doesn’t understand the context around the inquiries submitted.
During the presentation, Ramakrishan also said the organization was finishing up a pilot that would better structure its data within AI models so the technology can understand context around sales, inventory orders, traffic and corporate hierarchy.
“The goal is you can ask anybody in the company, 'What is the sales of this product last week?' or 'How many people visited my Soho store for a Coach purse last week?' or 'Compare the sales of Chicago and New York for women's bags.' So you can ask questions in any combination and it would answer,” he said.
