Databricks Generative AI Engineer Associate Question 126
Select 3You are building a customer support chatbot using a generative AI model on Databricks. The chatbot must answer user queries by leveraging additional context extracted from the user's input, such as product names and issue descriptions, to generate personalized responses. Which of the following approaches will help you augment the prompt with relevant context effectively?
- A
Extract key terms like product names and error codes from the user's input using a Named Entity Recognition (NER) model, and append them to the prompt for the generative AI model.
- B
Ignore the user's input and rely on pre-trained prompts to handle all potential customer queries.
- C
Use a keyword-matching algorithm to detect intents from the input and include the identified intent in the prompt for the generative AI model.
- D
Incorporate additional user-provided metadata, such as account type, into the prompt to refine the generative AI model's outputs.
- E
Allow the generative AI model to infer all necessary details without explicitly augmenting the prompt with user input.
Show answer and explanation
Correct answers: A, C, D
Explanation
To effectively augment a generative AI model's prompt, it is essential to extract and include relevant context from the user's input, such as key terms, intents, and metadata. This ensures that the model has the necessary information to generate accurate and personalized responses. Techniques such as NER, intent detection, and metadata incorporation are critical for achieving this, while ignoring user input or relying solely on the model's inference capabilities leads to suboptimal results.
- A. Correct.
Correct: Using an NER model to extract key terms like product names and error codes helps provide specific context to the generative AI model, leading to more accurate and personalized responses.
- B. Incorrect.
Incorrect: Ignoring the user's input and relying solely on pre-trained prompts will likely result in generic or irrelevant responses, as the model lacks the specific context required for personalization.
- C. Correct.
Correct: Using a keyword-matching algorithm to detect intents ensures that the generative AI model understands the user's purpose and can generate responses aligned with the user's needs.
- D. Correct.
Correct: Incorporating additional metadata, such as account type, into the prompt provides further context that can refine the AI model's output to better suit the user’s situation.
- E. Incorrect.
Incorrect: Relying on the generative AI model to infer all details without explicit context augmentation risks misinterpretation, as the model works best with clear and structured information.