Amazon Bedrock Knowledge Bases now supports metadata filtering to improve retrieval accuracy

This blog post announces that Knowledge Bases for Amazon Bedrock now supports metadata filtering to improve retrieval accuracy when using foundation models for Retrieval Augmented Generation (RAG).

Specifically, the article covers:

• An overview of metadata filtering and how it improves accuracy and performance by allowing pre-filtering of retrieval results based on custom metadata.
• A solution overview with steps to prepare a dataset with metadata, create a Knowledge Base, ingest data, and query with metadata filtering using the AWS console or SDK.
• Examples of common use cases for metadata filtering, such as document chatbots, conversational search, and intelligent search for developers.
• Instructions to clean up the resources created for the demo.