Genesys Cloud - Main

Hi everyone,

I've recently started encountering an issue with the "Get Participant Data" step in my Dialog Bot Flow. Occasionally, it fails to retrieve the participant information, yet the bot continues to the next step as if it succeeded. This causes problems because the bot variables remain set to NOT_SET, which affects downstream logic.

To mitigate timing issues, I've added a 1-second wait before calling "Get Participant Data"-assuming it would give enough time for the participant data to be available after transitioning from the inbound call flow. However, the issue still persists intermittently.

Has anyone else experienced similar behavior? Any guidance or best practices on how to resolve or work around this would be greatly appreciated.

Thanks in advance!

Hello Harisankar, 

I think there are several points we can look at. First being the difference between dialog bot flow and voice flows. Unlike voice flows the bot flows handle participant data differently, they don't maintain a queue for updates but instead directly fetch data from the cloud. This direct fetching mechanism can sometimes lead to timing issues. 

There is also some limitations to consider. Theres a limit of 50 Get Participant action executions per bot sessions. If the limit is reached the session will close and follow the configured error handling path. 

Some ideas to troubleshoot this are proper error handling, instead of just adding a wait step, implement proper error handling paths, add validation checks after the Get Participant data step, and use conditional logic to verify if the required variables are properly set before proceeding. 

You can also try variable validation by adding a decision step after Get participant data to check if the variables are NOT_SET and then create an error handling path that either retries the Get Participant Data or handles the error gracefully. 

The 1 second wait you've implemented might not be sufficient because of the timing of data availability can vary instead of a fixed wait it might be better to implement a proper validation and retry mechanism. 

Hope this helps!