Q5Operational Efficiency and Optimization for GenAI Applications
An enterprise application uses an Amazon Bedrock foundation model (FM) to process and analyze technical documents of 50 to 200 pages. Users experience inconsistent responses and truncated outputs when processing documents that exceed the FM's context-window limits. Which solution will solve this problem?
← → navigate · a answer
Community votes
Discussion · 7
C 3
Amazon Bedrock Knowledge Bases supports semantic chunking, which splits text based on meaning rather than fixed size. That helps improve retrieval quality when users ask about specific parts of technical documents. Then RetrieveAndGenerate retrieves only the most relevant chunks for the query instead of trying to fit large sections into the model context window, which helps avoid truncation and inconsistent
C 3
C should be
C 2
Semantic Chunking uses an embedding model to "look" at the meaning of sentences. It only creates a "breakpoint" when the semantic meaning changes significantly (controlled by the percentile threshold). The buffer size ensures that surrounding sentences are kept together, preserving the technical context necessary for an accurate summary.
B 2
Option B provides:
Managed chunking
Managed retrieval
Parent-child contextual expansion
Minimal operational overhead
This is the most robust and AWS-native approach.
D 1
I'm thinking D. Can anyone confirm this?
B 1
Child chunks ensure fine‑grained relevance where as Parent chunks ensure the model receives sufficient context without exceeding limits.
https://docs.aws.amazon.com/bedrock/latest/userguide/kb-chunking.html
1
Changing to C.