Amazon ElastiCache now supports real-time hybrid search with vector and full-text

This article announces that Amazon ElastiCache now supports real-time hybrid search combining vector similarity with full-text search in a single query.

• Combines semantic meaning with exact keyword matching for more relevant results
• Supports billions of embeddings from Bedrock, SageMaker, Anthropic, and OpenAI
• Delivers microsecond latency with up to 99% recall
• Data searchable immediately after writes complete
• Enables AI agent memory and RAG systems with improved context retrieval
• Available in all commercial AWS Regions, GovCloud, and China Regions
• Requires ElastiCache version 9.0 for Valkey at no additional cost

ElastiCache hybrid search enables developers to build more intelligent search experiences combining semantic and keyword matching without separate search infrastructure.