QuestionQ92

VerifiedOperational Efficiency and Optimization for GenAI Applications

A company uses Amazon Bedrock to build a Retrieval Augmented Generation (RAG) system. The RAG system uses an Amazon Bedrock knowledge base that relies on an Amazon S3 bucket as its data source for emergency-news video content. The system retrieves transcripts, archived reports, and related documents from the S3 bucket.

The RAG system uses state-of-the-art embedding models and a high-performing retrieval configuration. However, users report slow responses and irrelevant results, resulting in lower user satisfaction. The company observes that vector searches evaluate too many documents across too many content types and over excessively long time periods.

The company determines that the underlying models would not benefit from additional fine-tuning. The company must improve retrieval accuracy by applying smarter constraints. The company wants a solution that requires minimal changes to the existing architecture.

Which solution meets these requirements?

Explanation

Amazon Bedrock Knowledge Bases can associate metadata attributes with documents in an Amazon S3 data source and apply metadata filters during retrieval. Indexing metadata such as content type and date enables retrieval to restrict the candidate set, improving relevance and reducing unnecessary vector-search evaluation without replacing the existing knowledge base architecture.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!