問題文
A support team answers about 4,000 emails a month. Roughly 70 percent of them are variations of the same 20 questions, and the answers are documented in an internal knowledge base. The team wants to cut the time spent on first drafts. Which approach best fits this need?
選択肢
- Train a classification model that routes each email to the correct team member.
- Use a generative AI solution that composes replies from the model's general knowledge and sends them automatically, since the 20 questions are common ones.
- Use a generative AI solution that drafts replies grounded in the internal knowledge base, with an agent reviewing and sending each reply.
- Replace the knowledge base with a set of fixed email templates that agents copy manually.