Innovating with generative AI across Amazon

Amazon's AI Shopping Guides uses Amazon Bedrock to consolidate key product information and recommendations across over 100 product types to help simplify product research and discovery for customers. These guides, available in Amazon’s U.S. app (iOS and Android) and mobile website, help customers quickly get up to speed on products they might be less familiar with, reducing the time spent researching before making a purchase.

Amazon One Medical is using AI-powered tools to simplify administrative tasks for doctors, reducing their workload by 40% and enabling them to spend more time building relationships and providing high-quality care for their patients. Using AWS HealthScribe and Amazon Bedrock, Amazon One Medical developed solutions that automate note-taking, summarize medical histories, enable responsive patient communication, and streamline care team collaboration, allowing providers to focus on delivering exceptional, personalized care.

Amazon Ads expands its AI-powered creative tools with Audio Generator, joining Image Generator and Video Generator. These tools help brands quickly build and optimize campaign creatives using just product page info, enabling more experimentation. Data shows brands using Image generator saw ~5% more sales on average. The AI-powered suite empowers brands to connect with customers across the shopping journey through personalized, high-performing creatives generated with minimal effort.

Rufus, Amazon's new generative AI-powered conversational shopping assistant, scaled with over 80,000 AWS Inferentia and AWS Trainium chips for Prime Day. With these chips, Rufus reduced its costs by 4.5 times lower than other evaluated solutions while maintaining low latency for its customers.