A company has a Retrieval Augmented Generation (RAG) application that uses a vector database to store embeddings of documents. The company must migrate the application to AWS and must implement a solu

Question

Options

• AUse an AWS Batch job to process the files and generate embeddings. Use AWS Glue to store the
• BUse a custom Amazon SageMaker notebook to run a custom script to generate embeddings. Use
• CUse the Amazon Kendra S3 connector to ingest the documents from the S3 bucket into Amazon
• DUse an Amazon Textract asynchronous job to ingest the documents from the S3 bucket. Query

• A
  3% (1)
• B
  17% (6)
• C
  72% (26)
• D
  8% (3)

Explanation

Amazon Kendra is purpose-built for intelligent, semantic search using natural language understanding, and its native S3 connector directly ingests documents from the already-migrated S3 bucket with minimal configuration - making it the most complete, managed solution for this use case.

Why the distractors fail:

• A (AWS Batch + AWS Glue): Glue is an ETL service, not a vector store or semantic search engine - it can transform data but cannot serve embedding-based queries.
• B (SageMaker notebook): Notebooks are a development/experimentation tool, not production infrastructure; generating embeddings in a notebook still leaves the hard problem of storing and querying them unsolved.
• D (Amazon Textract): Textract is an OCR/document extraction service for pulling text from images and PDFs - it has no semantic search capability and is the wrong layer for this problem.

Memory tip: Associate Kendra = "Knowledge Search" - when an AWS question asks for semantic or intelligent natural language search over enterprise documents, Kendra is almost always the answer because it's AWS's only fully managed service designed specifically for that problem, complete with pre-built connectors for sources like S3.

No community discussion yet for this question.