Amazon Bedrock AgentCore Memory now supports direct ingestion to long-term memory
Amazon Bedrock AgentCore Memory introduces the IngestData API, allowing developers to directly submit content for long-term memory extraction without first persisting it as a short-term memory event. This change enables independent adoption of long-term memory, supporting both conversational and JSON payloads. Developers can verify extraction results, stream notifications via Kinesis, and redrive failed extractions. The functionality is available in all AWS Regions where Amazon Bedrock AgentCore Memory is supported.
Features (1) ›
- Direct content ingestion to long-term memory
Amazon Bedrock AgentCore Memory now provides the IngestData API for direct submission of content to long-term memory strategies, bypassing the previous requirement of storing content as a short-term memory event. This new capability supports both conversational payloads with USER/ASSISTANT roles and generic JSON payloads like behavioral or system events, along with optional metadata.
https://aws.amazon.com/about-aws/whats-new/2026/09/agentcore-memory-direct-ingest
Related releases
- How Sony LIV built real-time video streaming analytics with AWS AWS Big Data Blog ·
- Amazon Kinesis Data Streams now supports dry run for API request validation AWS What's New ·
- Kinesis Data Streams now supports Apache Iceberg streaming tables AWS Big Data Blog ·
- Amazon Kinesis Data Streams now delivers to general purpose S3 buckets AWS What's New ·
- Amazon Kinesis Data Streams introduces streaming tables for Apache Iceberg on S3 AWS What's New ·
- Amazon Redshift now supports 10MiB record ingestion from Kinesis Data Streams AWS What's New ·