AI-Powered Shopping tools

Shop-the-Show is Amazon's latest fan-fueled shopping experience using a Second-Screen. This delivers a seamless customer-first experience that allows viewers to simultaneously watch and shop their favorite shows, movies, and live sports events without interruption. The second-screen mobile experience is built around natural fan behavior, enabling customers to discover toys, apparel, and other products related to what they're watching when inspiration strikes, without disrupting the content they love.

The Shop-the-Show companion experience is built with AWS services including AWS AppSync to keep the dual-screen synchronization running smoothly and AWS Glue to prepare and integrate all the data. Scene detection is handled through the Prime Video X-Ray system built with Amazon Bedrock. Shop-the-Scene gets the frames from X-ray and uses them as ingress points to let customers open the Amazon Lens AI flow for product matching.

Rufus is designed to help customers save time and make more informed purchase decisions by answering questions on a variety of shopping needs and products. Rufus is an expert shopping assistant trained on Amazon’s product catalog and information from across the web to answer customer questions on shopping needs, products, and comparisons, make recommendations based on this context, and facilitate product discovery in the same Amazon shopping experience customers use regularly.

Rufus leverages AWS infrastructure to deliver its AI-powered shopping assistant at massive scale. It uses Amazon Bedrock to provide access to multiple foundation models including Amazon Nova and Anthropic Claude Sonnet, alongside Rufus's custom in-house model. This multi-model approach allows Rufus to optimize for different query types, balancing answer quality, latency, and cost. Rufus also uses Amazon Nova Web Grounding to retrieve and cite authoritative internet sources, improving accuracy and customer trust. The system integrates with various AWS services through tool calling capabilities, enabling real-time access to product catalogs, pricing databases, order history, and inventory systems. Amazon Bedrock managed infrastructure handles global scaling, model hosting, and deployment across multiple geographic regions, while features like prompt caching and parallel tool calling optimize latency.

Learn more: How Rufus scales conversational shopping experiences to millions of Amazon customers

View in your room lets customers view products in their home before they buy them. Customers can see how products fit and look before bringing them home. Curious whether a chair is the right size for your living room, or whether a coffee table you like will go along with your sofa? Now just pull out your phone to find out.

View in Room uses AWS services to enable customers to visualize products in their home before purchasing: Amazon S3 stores 3D models and product images, Amazon SageMaker Ground Truth labels training data for accurate product recognition and spatial positioning, and Amazon SageMaker develops and deploys the machine learning models that power real-time AR experiences at scale.