Resume Prompts are used when the end-user of the conversational AI application returns to a Flow after interrupting that Flow to discuss something else.
The purpose of Resume Prompts is to remind end-users of what they were discussing with the AI application previously so that they can resume that discussion or dialogue.
Imagine that the end-user is talking to the AI application about booking a hotel room. However, in the middle of the process, the end-user asks about the facilities available at the hotel.
When this happens, a different Flow (about the facilities) will be activated. And once this Flow has been completed, the application will resume the interrupted Flow (booking a hotel room), adding a text to remind the user that they were talking about booking a hotel room.
This text would be a Resume Prompt, for example: I thought we were talking about ...
In Teneo Studio, Resume Prompts can be added at two levels:
- to specific Output nodes, and
- to Flows.
When a user returns to a Flow, the more specific Resume Prompt (Resume Prompts in Output nodes) will be given first. If no Resume Prompts are defined at the Output node, the Flow Resume Prompt will be given.
If no Resume Prompt is specified in the Output node returned to, nor in the Flow that is returned to, the answer displayed will simply be the answer text of the Output node.
In order to resume a Flow, the Output node that was active when the Flow was interrupted must be set to Allow output to be revisited.
Allow output to be revisited is by default set to Limit number of revisits to 2
When an Output node doesn't allow to be revisited and the user input does not match the transition leaving that Output node, the Flow will be dropped and another Flow matching the user input will be triggered. When that Flow completes, execution will not return to the non-revisitable Output node. So, non-revisitable Output nodes behave like end nodes.
However, when the Output node does allow to be revisited and the user says something which at that point in the Flow is not recognized, another Flow will be triggered; and when the second Flow completes, the execution will return to the revisitable Output node.
Furthermore, it is possible to set a maximum limit for the number of revisits in order to allow the Output node to only be revisited a certain amount of times.