Amazon Bedrock Knowledge Bases now supports RAG evaluation (Preview)
News
Amazon Bedrock Knowledge Bases now offers RAG (Retrieval-Augmented Generation) evaluation capabilities in preview, providing a comprehensive tool for assessing AI application performance.
- Supports evaluation of information retrieval and generation processes
- Uses LLM-as-a-Judge technology with multiple judge models
- Retrieval metrics include context relevance and coverage
- Generation metrics cover correctness, completeness, faithfulness, and responsible AI indicators
- Allows comparison of Knowledge Bases with different configurations
The new tool simplifies RAG application testing, saving time and cost compared to manual evaluation, and enables faster application improvements directly within Amazon Bedrock.
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
Jul 10
2024
2024
Knowledge Bases for Amazon Bedrock now supports advanced RAG capabilities
Mar 20
2025
2025
Amazon Bedrock now supports RAG Evaluation (generally available)
Nov 22
2024
2024
Amazon Bedrock Knowledge Bases now supports binary vector embeddings to build RAG applications
Mar 14
2025
2025
Evaluating RAG applications with Amazon Bedrock knowledge base evaluation
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.