AWS launches Agentic Shopping Assistant to help retailers build AI tools

Amazon Web Services (AWS) has introduced the Agentic Shopping Assistant (ASA), a new tool designed to help retailers incorporate AI features into their online platforms. ASA combines AWS’s AI products, including Amazon Bedrock and OpenSearch, with professional services to streamline development. The service builds on Amazon’s Alexa for Shopping tool, which was added to its e-commerce marketplace this month and reportedly drove nearly $12 billion in incremental sales last year. ASA utilizes Amazon Bedrock, which provides access to cloud-based foundation models, and Bedrock AgentCore, an AI agent development toolkit. OpenSearch, an open-source search engine, enables AI assistants to quickly retrieve product information for users. AWS will offer professional services through systems integrators and the AWS Generative AI Innovation Center to tailor the tool to retailers’ needs, including adapting to brand guidelines and preventing irrelevant responses. The Generative AI Innovation Center will also develop custom workflows, such as personalizing responses based on shopper history, and analytics tools to monitor chatbot performance. AWS claims ASA can reduce the time to build custom AI applications from years to approximately 60 days. This launch follows Amazon’s recent moves to commercialize its e-commerce infrastructure, including a logistics service for businesses and potential robot sales to industrial companies. The service aims to address challenges retailers face in integrating AI, such as maintaining brand consistency and ensuring accurate, relevant outputs. By offering pre-built AI capabilities, AWS seeks to lower the barrier for smaller retailers to adopt advanced shopping assistants. ASA’s introduction underscores AWS’s broader strategy of repurposing its internal AI tools for external use, reinforcing its position in cloud-based retail innovation.