The Chatbot Connector Preview is deprecated effective July 1st ‘21
The service is deprecated and will be unavailable after October 2021. Learn more
- Guide to building your first Dialogflow NLP agent: https://codelabs.developers.google.com/codelabs/chatbots-dialogflow-appointment-scheduler/#0
- A sample chatbot for retail:
- Discussions and tutorials on the Chatbot topic from the UiPath Forum:
Can I use processes and unattended robots from multiple folders in Orchestrator?
You can only map intents to processes in a single folder in the connected Orchestrator service.
Can I map a single intent to multiple processes in the same connection?
It is not possible to map a single intent to multiple processes in the same connection. If your logic is spread across multiple processes, we recommend using the Start Job activity to start additional process runs.
When does a process start the execution?
A mapped process is scheduled for execution in Orchestrator as soon as the mapped intent is resolved, i.e., all required parameters have values provided by the end-user. The process begins the execution when its job is picked up by an available unattended robot.
Can I change the message sent to the user when a process is scheduled to run?
The Working on it... message sent to the user when a process is triggered by an intent cannot be configured at this time.
Can I change the timeout limit and message?
The timeout limit (5 minutes) and message cannot be customized.
I have a process that might take longer than 5 minutes to run. What do I do?
We recommend splitting up your process as follows:
1. The first process is triggered by the intent, stores data in Orchestrator queues, and sets the output message to respond to the user.
2. The second process picks up the queue item to perform time-intensive automation and send results to the user via email or Slack.
Can I create branching conversations?
Yes, you can! To conditionally respond to a user from a process, simply set the output argument of the process to a message that you'd like to display to the user (based on certain conditions). Next, define intents in Dialogflow for responses that users might have based on the message they receive from the chatbot.
How do I use information from previously triggered process runs later in the chat?
Any information that might be useful later in a chat conversation can be stored in Orchestrator queues for quick access using parameters already captured as input from the user.
Can I use voice with my chatbot?
The chatbot connector doesn't provide voice integrations at this time.
Can I integrate custom channels and NLP?
Custom channels and NLP cannot be integrated with the chatbot connector.
Where is my chat data stored?
The chatbot connector does not store any conversation data. All messages sent by the user are stored in, and accessible from Google Dialogflow per its policies. Any information retrieved from the RPA process is not sent to Dialogflow.
Can I deploy chatbots on-premise?
The chatbot connector is a cloud-only service and cannot be deployed on-premise. However, it can connect to an on-premise Orchestrator installation.