Answer: Create an Amazon Bedrock knowledge base.

## Explanation **Correct Answer: C. Create an Amazon Bedrock knowledge base.** ### Why this is correct: 1. **Amazon Bedrock Knowledge Base** is specifically designed to allow foundation models to access and use private, company-specific data sources. It enables Retrieval Augmented Generation (RAG) where the model can retrieve relevant information from your private data sources to supplement its responses. 2. **How it works**: - You create a knowledge base that connects to your private data sources (Amazon S3, databases, etc.) - The data is processed and indexed for efficient retrieval - When the model generates responses, it can retrieve relevant information from your knowledge base - This allows the model to provide answers based on your company's specific data ### Why other options are incorrect: **A. Use a different FM.** - Changing to a different foundation model doesn't solve the problem of accessing private company data. The issue is about data access, not model capability. **B. Choose a lower temperature value.** - Temperature controls the randomness/creativity of model responses (lower = more deterministic, higher = more creative). This doesn't help the model access private company data. **D. Enable model invocation logging.** - Model invocation logging is for monitoring and auditing purposes - it tracks API calls and usage but doesn't provide the model with access to private data sources. ### Key Concept: **Retrieval Augmented Generation (RAG)** is the technique used here, where foundation models are enhanced with external knowledge sources. Amazon Bedrock's Knowledge Base feature implements RAG by allowing models to retrieve information from your private data stores before generating responses.

Author: Ritesh Yadav