AWS Certified Cloud Practitioner

Explanation

The correct answer is B) Context enrichment.

Context enrichment refers to the practice of providing additional relevant information or context to an AI model's prompts to improve the quality and accuracy of its responses. In this scenario:

• The AI-powered FAQ bot was producing inconsistent answers to policy-related questions

• The developer added excerpts from the company's policy document to the prompts

• This provides the AI model with specific, relevant context about the company's actual policies

• With this enriched context, the AI can generate more accurate and consistent responses that align with the company's official policies

Why the other options are incorrect:

• A) Zero-shot prompting: This refers to asking a model to perform a task without providing any examples, which is not what's happening here since the developer is adding policy excerpts.

• C) Negative prompting: This involves telling the model what NOT to do, rather than providing positive context.

• D) Temperature adjustment: This refers to adjusting the randomness/creativity of AI responses, not adding contextual information.

Context enrichment is a common best practice in AI prompt engineering to ensure models have the necessary background information to generate appropriate responses.