So I’ve created a couple of chat agents with data stores. Initial testing is looking promising.
i have another set of data in an sql db that I can easily convert into jsonl and load into a data store. I could easily create another agent on that data store, but the requirement I have is to automatically produce a summary of this data on a regular basis - without user input (ie. the user doesn’t want to ask the agent for it, they just want to see the summary on an existing page when they go to it.
how would I do this?
I’ve seen what I think is the approach - write a python script/app to load a model and run the prompt. Is this what I need to do and if so how do i specify the agent and/or data source, and how is the result (text) output?
Hello TonyTheBubble,
According to the image below, only Dialogflow CX has the capability to do custom prompts for data store query rewrite and summarization.
Alternatively, you may use agent assist to generate conversation summaries. Here are the documentation for enabling agent assist custom model for feature lets you provide conversation summaries to your agents after each conversation is completed.
I hope the above information is helpful.