Routing email to queues

This video shows you how to create a new email topic and route email to a particular work queue. It also explains how text analytics is used for NLP (Natural Language Processing) topic detection and entity extraction.

U+, a telco, recently introduced an email channel for customer interactions. Currently, when customers email the contact center, the emails are routed to a common work queue and are not categorized. A work queue is a container that holds work that is waiting to be assigned to operators.

As a system architect, you have been tasked with building intelligent email routing into the system.

You use the Web Mail client to send two emails to the support center. The first email requests a change of address. The second email makes a complaint about how a recent payment by the customer was handled.

Each customer email generates a case in the Interaction Portal, the support center application. A case is a work request that is created in the application.

As CSR, you log in to the application and view open emails. Here you can see that the case is under a common work queue named Incoming emails, which routes email directly to an operator.

Click on the case that was created for the address change email. Notice that the case type is identified, but the mail is not routed to the right work queue.

View the complaint email. Notice that a case type or work queue is not identified.

To route emails to work queues, you configure the email channel interface with case types and configure routing of emails based on associated topics.

As administrator, you log in to App Studio. You click Channels, then click My Support. This is the email channel used by U+.

To configure a new case type, click the Behavior tab. The existing case types are listed under the Suggested Cases section. You can add as many case types as you need. You can see that the Address change case is already configured. You can view the keywords that, when found in an email, trigger the Address change case. In addition to the keywords, each case has an associated NLP model that learns to identify the email topic from its content.

For now, you are only interested in adding the File a complaint case to identify customer complaints that are sent via email.

You add the suggested case named File a complaint. In the Text analysis tab, you enter the words that are likely to trigger this case type. For example, you could enter "complaint", "wrong", "moved", "angry", "upset" as approximate match words. These definitions represent the rules that will be used for topic detection.

You have now configured the new case type, which associates a complaint email with the File a complaint case.

Next, configure the routing to ensure the emails are sent to the right work queues. Address change emails need to be routed to the Account Maintenance work queue. Complaint emails should route to the Complaints queue.

To route an email to a work queue, click on +Add condition. In the new condition configuration area, select the action, Route to a work queue and specify a work queue. In this case, Account Maintenance. Then, add a When condition with the NLP->Topic as Address Change. The Topic is the result of the NLP text processing that is executed on the email.

Add a condition to route to work queue Complaints, when topic equals File a complaint.

You have completed all the configuration steps. You can now save your configuration and start the test execution. Use the Test window to check your work.

Notice that the complaint email is now associated with the File a complaint topic and the outcome (the work queue) is Complaints.

Notice that the email is now associated with the Address Change topic, and the outcome (the work queue) is AccountMaintenance. You can also see how NLP has identified different entities in the text.

This video has concluded. What did it show you?

• How to create a case type in an email channel
• How to route an email to a desired work queue
• How text analytics is used for NLP-based topic detection and entity extraction