Databricks Generative AI Engineer Associate Question 152
Single answerYou are designing a customer support chatbot for a retail company. The chatbot must handle a wide variety of questions about products, orders, and returns. It needs to provide accurate, context-aware responses, support multiple languages, and prioritize cost-efficiency. Which LLM would be the best choice for this application?
- A
An open-source LLM fine-tuned on customer support datasets
- B
A general-purpose LLM hosted and maintained by a cloud provider with multilingual support
- C
A task-specific LLM pre-trained for e-commerce applications
- D
A lightweight LLM optimized for low-latency inference but with limited multilingual capabilities
Show answer and explanation
Correct answer: B
Explanation
For a customer support chatbot with requirements for multilingual support, accuracy, and cost-efficiency, a general-purpose LLM hosted by a cloud provider is the best choice. It provides the necessary capabilities out-of-the-box, including multilingual support and scalability, while reducing operational costs and complexity compared to fine-tuning or hosting models independently.
- A. Incorrect.
While an open-source LLM fine-tuned on customer support datasets can provide some customization, it might require significant engineering effort for multilingual support and may not be cost-effective in terms of infrastructure requirements.
- B. Correct.
A general-purpose LLM hosted and maintained by a cloud provider with multilingual support would meet the requirements for accuracy, context-awareness, multilingual capabilities, and cost-efficiency, as it eliminates the need for extensive infrastructure and maintenance.
- C. Incorrect.
A task-specific LLM pre-trained for e-commerce applications might be useful for domain-specific nuances but may not adequately handle multilingual support or general customer queries beyond e-commerce contexts.
- D. Incorrect.
A lightweight LLM optimized for low-latency inference would be cost-efficient but likely fails to meet the accuracy and multilingual requirements for the chatbot.