Question 22
Domain 2: Fundamentals of Generative AIWhich AWS service or feature stores embeddings in a vector database for use with foundation models (FMs) and Retrieval Augmented Generation (RAG)?
Correct answer: B
Explanation
Amazon OpenSearch Service stores embeddings in a vector database because it supports “approximate nearest-neighbor vector search” and a “vector engine” for semantic search. The source says it is “the default vector backend” for Amazon Bedrock Knowledge Bases and is optimized for “large-scale, high-throughput” RAG workloads.
Why each option is right or wrong
A. Amazon SageMaker Ground Truth
Amazon SageMaker Ground Truth labels data for training, not vector storage for retrieval.
B. Amazon OpenSearch Service
Amazon OpenSearch Service is the vector-store choice because the question asks for the AWS service that stores embeddings in a vector database for RAG, and OpenSearch supports approximate nearest-neighbor vector search through its k-NN capability and vector engine. The cited material states it is optimized for large-scale, high-throughput semantic search and is the default vector backend for Amazon Bedrock Knowledge Bases when no alternative is specified, which matches the RAG use case described here.
C. Amazon Transcribe
Amazon Transcribe converts speech to text; it does not store embeddings or run vector search.
D. Amazon Textract
Amazon Textract extracts text and structured data from documents; it is not a vector database.