Deliver hyper-personalized viewer experiences with an agentic AI movie assistant using Amazon Bedrock AgentCore and Amazon Nova Sonic 2.0

This article demonstrates building a conversational AI movie assistant using Amazon Bedrock AgentCore and Amazon Nova Sonic 2.0 for personalized viewer experiences.

• Agentic AI combines ML pattern recognition with generative AI's contextual understanding for better recommendations
• Natural speech interface using Nova Sonic 2.0 enables real-time, human-like voice conversations with low latency
• Movie recommendation flow uses intent classification, query rewriting, semantic search, and reranking for top results
• Scene analysis feature provides summaries and actor identification using Amazon Bedrock Data Automation
• Architecture uses WebSocket connections, AWS Fargate, Lambda, OpenSearch, and S3 Vector for semantic search
• System maintains user affinity profiles in DynamoDB to personalize recommendations based on viewing history
• Multi-turn conversations preserve context throughout sessions for enriched user experience

The solution transforms movie discovery from implicit feedback to explicit conversational preference gathering, improving engagement and retention through voice-driven personalization.