Amazon Web Services has launched Agentic Shopping Assistant, or ASA, a new offering that helps retailers build AI shopping features into their online stores. ASA bundles several AWS products with professional services, drawing on Amazon Bedrock and its AgentCore agent-development toolkit for the underlying foundation models, and using a managed version of the OpenSearch search engine to retrieve product information. The service is derived from the Alexa for Shopping tool that Amazon added to its own marketplace this month, which replaced a set of AI features that Amazon said drove nearly $12B in incremental sales last year. Through its Generative AI Innovation Center and outside systems integrators, AWS will help retailers customize the assistant to their brand voice, design guidelines, and product catalog, add guardrails against irrelevant or inaccurate output, build chat-history personalization, and add analytics to monitor chatbot quality. AWS says building a custom AI application from scratch can take years, while ASA compresses the process to roughly 60 days.
