Question 34
Domain 3: Applications of Foundation ModelsA company wants to create a chatbot to answer employee questions about company policies. Company policies are updated frequently. The chatbot must reflect the changes in near real time. The company wants to choose a large language model (LLM). Which solution meets these requirements?
Correct answer: C
Explanation
RAG fits because the chatbot can retrieve the latest policy documents at query time, so it reflects frequent updates without retraining the LLM. Amazon Bedrock Knowledge supports this retrieval layer, matching the need for “near real time” policy changes while using an LLM for generation.
Why each option is right or wrong
A. Fine-tune an LLM on the company policy text by using Amazon SageMaker.
SageMaker is commonly used for training jobs, not for keeping answers updated in real time.
B. Select a foundation model (FM) from Amazon Bedrock to build an application.
A foundation model is the base LLM; it does not by itself add live document retrieval.
C. Create a Retrieval Augmented Generation (RAG) workflow by using Amazon Bedrock Knowledge
Amazon Bedrock Knowledge is the retrieval layer that lets the chatbot pull the current policy documents at request time, so the answer reflects updates immediately instead of waiting for model retraining. That matters here because LLM training is a separate, expensive process on services like Amazon SageMaker AI and is not the mechanism for near-real-time content changes; the model can stay fixed while the knowledge base is refreshed as policies change.
D. Use Amazon Q Business to build a custom Q App.
Amazon Q Business is for building Q Apps, not the general RAG workflow described here.