Trigger robots from Rasa chatbots
Example robot that returns current time in specified city to RASA. Demonstrates two-way integration between chatbots and RPA.
Chatbots and RPA robots are a great match creating great business outcomes together. RPA robots extend the capabilities of conversational user interface to be able to perform tasks with legacy systems that lack simple API connectivity. This robot is made as a part of the solution architecture to demonstrate end-to-end capabilities of platforms when integrated together.
What you'll learn with this robot
- Using Control Room API to trigger robots with a work item from chatbot conversations
- Using Basic Selenium browser automation with Google.com
- Using Work Data Management to create work items between robot steps
RPA.HTTPto call a webhook to return data to Rasa conversation
Example consists of two robots: one that gets the current time from Google using browser automation, and one that sends the result back to Rasa chatbot conversation.
This example uses the
rasa.robotfile to return results the the chat conversation. An additional
email.robotis included should you want to skip sending the results to Rasa, but want them send to email instead.
In order to set the demo up, please create a robot with two steps in the Control Room.
Also notice that in order for the robot to be able to return results back to Rasa conversations, the robot needs to run locally in the same environment as the chatbot server using Robocorp Workforce Agent, or otherwise have access to the RASA endpoint.
The exact steps required to configure the process:
- Have or create a Robocorp Control Room account.
- Install Workforce Agent
- Connect Workforce Agent with Control Room to create an execution environment.
- Create a robot from public git repository (this repo). Note: if you are running the Rasa service somewhere else than
localhost:5005you'll need to clone/fork the repository and make required changes.
- Create a new Process under Workforce tab
- Add Search Time as the first task, and choose the execution environment to be your newly linked local environment.
- Add Send to RASA as the second task, and choose the execution environment to be your newly linked local environment.
- Follow instructions on Rasa example repo how to get the Rasa server running.
:heart: Have fun chatting!