If I have a big list of ‘solution’ templates from which the customer support representative replies to a customer’s query, how do I go about feeding this set of templates as a model for the bot to reply and based on the initial Q&A AI generated responses, the human customer support starts responding ?
This is not a real-time chat but replying via a ticketing platform which has an API to auto-feed in replies.
So I was wondering how would I go implementing TensorFlow for model input and get generated responses.