I am trying to automate the user message input while running interactive learning. Idea is to create a script which will fetch the user messages with confidence less than some threshold value from the tracker-store (MongoDB in my case). Then pass those messages in the interactive learning process.
I’m currently looking for the way to override the _enter_user_message method in interactive.py with creating any patch in the rasa core itself.
Mhm, I think for a proof of concept, you could just overwrite that function - monkeypatching in your functionality, e.g. something along these lines should work: