What AWS’s Agentic Shopping Assistant Is and Why It Matters
AWS’s Agentic Shopping Assistant is an AI shopping assistant built on Alexa for Shopping that lets retailers create conversational, brand-specific online buying experiences without building their own retail AI technology from scratch. It gives shoppers a natural language interface that can answer questions, explain differences between products, and recommend items based on a retailer’s catalog and rules, shifting online commerce toward agentic commerce powered by autonomous AI agents that handle more of the shopping journey. Amazon says the same Alexa for Shopping infrastructure previously drove nearly $12 billion in incremental sales for its own marketplace, and Accenture estimates that more than 30% of online commerce could run through AI agents by 2030. By opening this proven stack to outside merchants through AWS, Amazon turns a once-exclusive advantage into a shared AI shopping assistant platform.
From Alexa for Shopping to Agentic Commerce as a Service
The AWS Agentic Shopping Assistant represents Amazon’s familiar playbook: productize internal tools for external customers. The system uses the Alexa for Shopping technology that underpins Amazon.com’s AI assistant, previously known as Rufus, and reconfigures it so retailers can deploy AI shopping assistants tuned to their own inventory, tone, and policies. AWS says retailers can launch a working agentic commerce experience in around 60 days, suggesting a relatively fast path from pilot to production. Amazon emphasizes that merchants keep control over customer data, catalogs, and business rules, even though the infrastructure runs on AWS while Amazon’s retail arm competes for many of the same shoppers. One early adopter, Kate Spade’s parent company Tapestry, built an AI gift concierge that asks about the gifting occasion, recipient, and style, then proposes products. Amazon notes this concierge uses Anthropic’s Haiku 4.5 model through Amazon Bedrock and went through about 2.5 months of testing.
How Agentic AI Could Rewrite Retail Competition
Agentic commerce shifts competition from static catalogs and search bars to AI shopping assistants that actively guide each purchase. Instead of users manually filtering products, conversational AI agents interpret intent, cross-reference stock, and refine options in real time. For Amazon, turning its Alexa for Shopping stack into a service broadens its influence across the retail industry, even when purchases happen off Amazon.com. For retailers, the decision is more complex: adopting AWS’s agentic commerce means relying on a rival’s infrastructure while gaining access to high-performance retail AI technology they might not build alone. The potential prize is large. According to Accenture, more than 30% of online commerce could run through AI agents by 2030, representing about $3.1 trillion in transactions. If that forecast proves accurate, early movers in AI shopping assistant deployment will likely shape customer habits and expectations for years.
Nosto, Shopify, and Multi-Channel Agentic Experiences
AWS’s move lands alongside parallel efforts in the Shopify ecosystem, where Nosto is bringing agentic commerce deeper into back-office workflows. Nosto’s latest release connects its Commerce Experience Platform with Shopify’s AI assistant, Sidekick, so merchandisers can steer product discovery and personalization using conversational instructions without leaving Shopify. At the OMR Festival, Ford’s ecommerce partner Autonative used Sidekick prompts to change the design of Nosto-powered recommendations on its online accessories store, reducing the need for developers and cutting execution overhead. Nosto is also building Huginn, an AI agent that links through MCP to expand agentic workflows, from documentation help via a Doc Agent to proactive merchandising insights and automated actions via a Merchandising Agent. With integrations like these, smaller retailers running on Shopify gain tools to compete with AI-driven experiences across channels, especially when paired with front-end AI shopping assistants powered by platforms such as AWS.
What Retailers Should Watch Next
Taken together, AWS’s Agentic Shopping Assistant and Nosto’s Shopify workflows mark a turning point: AI shopping assistants are moving from experimental features to standard infrastructure. Retailers now face a strategic choice between building proprietary agentic commerce stacks, adopting cloud platforms like AWS, or layering specialized tools such as Nosto on top of existing systems. Key questions include how much control they want over models and data, how deeply AI agents should automate merchandising decisions, and how to align AI recommendations with brand identity. Early examples, like Kate Spade’s AI gift concierge and Ford’s Sidekick-driven recommendations, show that agentic commerce can start with focused use cases and expand over time. As AI shopping assistant capabilities spread, differentiation will depend less on having retail AI technology at all and more on how uniquely each retailer trains, configures, and governs its agents.