Right now, the Reputation responses are generic, and not in the voice of the brand. It does not link to the knowledge base provided in the Conversation AI or anywhere else, so it can get the context of an answer completely wrong.

I'd like to be able to provide bot training on certain types of complaints, with custom responses and links. This will make it much easier to handle negative responses correctly.

Also, please stipulate the AI responses must be grammatically correct. There should not be tense or subject/verb agreement mistakes happening in the responses.