Configure the data flow that retrieves audio transcript data from Pega Cloud and processes the data.
To ensure proper functioning of Voice AI, you configure audio transcript retrieval data flow to run on the RealTime node, with a single thread retrieving audio transcripts from the Pega server.
Once you run the data flow, you can monitor the data flow status on the Real-time processing landing page or search for the run ID within the Pega-DM-DDF-Work class. You can stop or start the data flow as needed, but if the data flow is stopped, the Voice AI feature cannot work. For more information, see Data flow not running.
- In the header of Dev Studio, click .
- On the Real-time processing landing page, click New, and
then specify the following settings:
- In the Applies to list, select the Work-Channel-Interaction-VoiceAI class.
- In the Data flow list, select VoiceAIRetrieval.
- In the Access group list, select the current
implementation access group.This access group must include the VoiceAI:CSR role. For more information, see step 4 in Additional preparatory configuration.
- In the Service instance name list, select RealTime.
- In the Number of threads field, enter 1.
- To make it easier to find the data flow, select the
Manage the run and include it in the application
checkbox, and then complete the following fields:
- In the Ruleset list, select an open ruleset with which you want to associate the data flow.
- In the Run ID field, specify a unique name for the data flow. For example, VoiceDataRetrieval.
- Click Run.
Note: Some browser plugins might interfere with creation of the data flow work item. If an error is displayed that the system cannot find the VoiceAIRetrieval data flow, try disabling all extensions or creating the data flow in a private browser. Result: The data flow status is displayed as In progress, indicating that the data flow is in the process of running.