What are service request classifications?

When your customer sends a service email for a new issue, the email is converted to a Service Request (SR). You can quickly classify service requests by auto-populating the Product, and Category fields. Use the Service Request (SR) Classification feature, which applies to requests created by the inbound email channel, to streamline the overall process of categorizing and assigning SRs.

The request should have its Title, Problem Description, and Primary Point of Contact prepopulated. However, fields such as Category and Product that are useful in routing the request to its proper queue and assignee are blank. By enabling the Service Request (SR) Classification feature, the Product, and Category can be auto populated when the request is created. The agent doesn't need to populate it manually. This feature applies to requests created by the inbound email channel.

Note: This feature is currently available for English language only.

Attribute predictions come from the Adaptive Intelligence (AI) applications platform that's already provisioned for you.

Here’s a brief outline of how the SRs are classified:

1. The database contains records of all resolved requests created from past inbound emails. These requests have either product, category, or both already populated. Resolved requests are considered correct for the purpose of building the Machine Learning (ML) model within the AI application.
   Note: Accurate model predictions rest on the fact that historical resolved requests have correct labels.
2. You must activate the tenant from the Configure SR Classification Using Adaptive Intelligence page. After you activate the tenant, data from closed requests is imported by the AI application.
3. The AI model automatically learns, so it returns the predicted product or category based on data from closed requests in Fusion Service.
4. Based on analysis of the subject line and body of the email, the model classifies requests and returns products and categories that have the highest confidence.
5. While reviewing the assigned request, an agent can manually update the AI-labeled fields, if required.