Hi,
I am working with the idea of answering tickets in a few select categories or with specific keywords with OpenAI.
For example if a user submits a ticket saying a printer is not working, then OpenAI should add a reply to the ticket with instructiosn on how to troubleshoot. Like a supporter would. If the user replies back that it is not working, it goes into to the regular queue for a human to take care of it.
Since both SDP and OpenAI have API. I was wondering what the best approach to test this would be?