Introducing seven new features of the Agentic Shopping Assistant on AWS

This article introduces seven new features for the Agentic Shopping Assistant on AWS, a generative AI solution that helps retailers provide personalized shopping guidance similar to in-store associates.

• Self-service admin portal enables business users to manage brand voice, catalog updates, and external tools without development cycles
• Long-term memory remembers shopper preferences and prior conversations across sessions using Amazon Bedrock AgentCore
• Multimodal search allows shoppers to search by product image, voice, or text with voice-to-voice responses powered by Amazon Nova Sonic
• Conversation session history selector lets shoppers browse and revisit past conversations organized by auto-generated topic labels
• Cart awareness and management enables shoppers to view, add, remove items, and receive complementary suggestions without leaving the conversation
• Shopify integration connects existing catalogs through Universal Commerce Protocol with no custom API mapping required
• Conversational intelligence dashboards provide real-time demand signals, trending categories, and operational metrics for merchandising insights

These features enable retailers to deliver personalized, visual, and complete shopping experiences from discovery through checkout while maintaining data ownership and transparency.