Hello there,
I use Rasa for my final thesis and wanted to include a mechanism to let the chatbot provide intelligent answers to too many fallbacks. For this purpose I analyze the conversation up to this status and compare it with old conversations in my database. The results fit all so I get the most likely next action back. I want to send them prepared in a Rasa-readable format to the EventBroker, so that the predicted action is streamed to the EventBroker and executed.
How can I do this?
I use a postgres database as a docker container to persist everything. My entire chatbot runs in a docker-compose-environment.
My previous attempt is to simply add a new entry to the EventBroker. This works so the entry is inside the postgres database, but the action isn’t triggered. This looks like the following example:
{ “sender_id”:“62eba9f4-91a6-46da-bf5d-d82a8b03c5fe”, “event”:“action”, “timestamp”:1596795321.2934031, “name”:“action_study_course_search”, “policy”:“policy_1_TEDPolicy”, “confidence”:0.9716068506240845 }
I hope you can help me. If you need more information, please let me know.
Thanks a lot in advance!
LG
Xanenon