Profanity message filter overview

Last Updated : Dec 13, 2022 |
Prolog information
The profanity filter helps you block the sending and receiving of responses to customers and agents if it contains profanity words. It is important to immediately identify the sensitive words used in the conversations and notify if they contain profanities.
The profanity filter works based on the value of the FilterEnabled flag. The filtering service scans the message and masks any text it does not want the destination party to see.
Note:
The adopting client is responsible for implementing the filtering service.
When the profanity filter is enabled, all messages sent from the end user to the contact center or from the agent to the end-user result in a call to the configured REST service. The following fields form part of the request:
  • Message Text
  • Channel Source
  • Sender
  • Destination
The filter service responds within the timeout window, set by default to 5 seconds. If the filter service is enabled and no response is received from the service call within the configured timeout period, each message raises an alarm. The Time out filter ensures the administrator is notified that the filtering service is unreachable or slow in responding.
After the timeout period, the async message is processed as normal and the original message is sent to the end-user or agent as normal.
A workrequestID is generated and sent to the profanity filter for all messages sent from EndUser to Agent. The sendWorkRequestID parameter is added to the deployment.yaml file in AsyncMsgConnector to enable or disable this functionality.