Introducing Amazon OpenSearch Memory: Persistent memory for AI agents
Big Data Blog
This article announces the general availability of Amazon OpenSearch Memory, a serverless capability for adding persistent memory to AI agents built on Amazon OpenSearch Serverless.
- Stores agent interactions as searchable events and derives structured memories automatically
- Reduces input tokens by ~88% compared to full conversation history while maintaining 93% answer accuracy
- Handles extraction, embedding, indexing, consolidation, retrieval, and data isolation without infrastructure provisioning
- Organizes data in spaces containing events (original interactions) and memories (facts, preferences, summaries)
- Supports keyword, semantic, and hybrid search across memories and events
- Available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Frankfurt)
- Integrates with AWS SDK, Model Context Protocol server, and supported frameworks
OpenSearch Memory enables AI agents to maintain long-term context across sessions without building separate memory systems, reducing costs and improving agent responses.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Nov 16
2026
2026
Build memory-aware AI agents with Agentic Retrieve on Amazon Bedrock Managed Knowledge Base
Jun 8
2026
2026
Building persistent memory for multi-agent AI systems with Amazon S3 Vectors
May 28
2026
2026
Introducing the next generation of Amazon OpenSearch Serverless for building your agentic AI applications
Nov 17
2026
2026
Amazon Bedrock AgentCore Memory now supports short-term memory retention of up to 10 years
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.