NCA-GENL Question 74
Select 3A company is using a Large Language Model (LLM) to assist its sales team in making strategic decisions based on customer feedback. Which of the following practices would ensure the model outputs relevant and actionable insights for decision-making?
- A
Fine-tuning the LLM with domain-specific customer feedback data.
- B
Ensuring the model is regularly updated with the latest general pre-trained weights.
- C
Using prompt engineering to frame questions in a way that aligns with the sales team's goals.
- D
Integrating the LLM with external tools, such as CRM systems, for contextual data enrichment.
- E
Relying solely on the LLM's default capabilities without additional customization.
Show answer and explanation
Correct answers: A, C, D
Explanation
To support decision-making effectively, an LLM must be both tailored to the domain and integrated into the business workflow. Fine-tuning ensures the model understands the specific context, while prompt engineering and integration with external tools like CRM systems enhance its ability to deliver actionable and relevant insights. Relying only on default capabilities or generic updates is insufficient for driving strategic decisions in a specialized domain.
- A. Correct.
Fine-tuning the LLM with domain-specific customer feedback data ensures the model understands the unique language and context of the company's customers, making its insights more relevant.
- B. Incorrect.
While updating with the latest pre-trained weights can improve general performance, it does not address domain-specific needs or ensure actionable insights for the sales team.
- C. Correct.
Prompt engineering helps guide the LLM to generate responses that are aligned with the sales team's objectives, improving the relevance and usefulness of the outputs.
- D. Correct.
Integrating the LLM with tools like CRM systems allows the model to access contextual and up-to-date customer data, which enhances its ability to provide actionable insights.
- E. Incorrect.
Relying solely on the LLM's default capabilities without customization often results in generic outputs that may lack the specificity needed for informed decision-making.