I am trying to write my own agent which interacts with the nlu interpreter. The agent will follow a flowchart based approach. Is there any documentation on which methods my agent needs to implement. The agent should take instead of stories, a json flowchart of the conversational flow.
I’m afraid this isn’t how Rasa Core was designed to work, you’d have to convert that to the stories format. Are you by chance using botsociety? If you use it in beta mode, there’s an “export to Rasa” button, which would export your designed stories to the stories format
Hi @akelad, I have built a rasa core which is working great! But, when two different users enter the input at same time, the answers are supplied in reverse order. What is the best way, so that, rasa responds to each user, for their query. How can rasa agents be used in this scenario ?
@sarmaemani not sure if you’ve since resolved your issue, but each user should have a unique tracker id, then talking to the bot at the same time works fine