Configure AI topic clustering for cases
Important
Power Virtual Agents capabilities and features are now part of Microsoft Copilot Studio following significant investments in generative AI and enhanced integrations across Microsoft Copilot.
Some articles and screenshots might refer to Power Virtual Agents while we update documentation and training content.
Customer Service Insights uses AI to give you insights into your customer service data by grouping semantically related cases and generating a topic. New cases that match the generated topic are automatically added to the topic group. This action can help you identify areas for improvement that can have the greatest impact on system performance.
The AI-driven technology empowers informed decision-making about how to improve resolution rates, reduce wait times, and decrease customer service costs. You can use case resolution insights, backlog trends, and historical comparisons to evaluate agent performance and business impact, and address inefficiencies in your system.
Enable topic clustering for cases
Topic clustering is enabled by default when you enable historical analytics. To enable historical analytics, see Configure Customer Service historical analytics.
Summary and Model run summary views
The Summary and Model run summary views provide key information about how the topic model is operating.
| View | Description |
|---|---|
| Status | Whether the feature is on. |
| Data attributes used | Which text field from the Case entity is used for topic generation. |
| Topics generated | The total number of topics generated by the model. |
| Cases associated to a topic | The percentage of cases that were considered for topic generation and classified to a topic. |
| Last successful run | Timestamp of the last time new cases were processed. |
| Run frequency | The cadence in which new cases are processed and tagged with topics. |
Data mapping
Data mapping enables you to choose the text field that the agents in your organization are most likely to use to describe the reason why a customer reached out to support. By default, the Case Title attribute is used. The other available attributes are Serial Number, Description, and Case Title.
Improve data quality by cleaning support case data
The AI Insights charts that are displayed on the Customer Service Insights dashboards are generated by applying language understanding technology to the titles of support cases. However, the results can be misleading if the titles include extraneous information such as product name, case status, or ticket number tags. You can improve the quality of the results that are displayed in AI Insights charts by specifying Data Cleaning settings to disregard tags in titles when they're grouped into topics, and specific phrases that should be ignored. When you choose to apply both options, sections are ignored first, followed by phrases.
Enable topic automation for Copilot Studio
AI discovered topics in Customer Service Historical Analytics are often prime candidates as topics for automation in Copilot Studio bots. If Copilot Studio is available in the region that your Customer Service organization is in, you can enable the feature
- In Customer Service admin center, go to Insights > Topics clustering for cases > Manage.
- In the Topic automation section of the Topic clustering for case page, enable the toggle.
Note
Topic automation to Copilot Studio bot is currently not supported in Government Community Cloud.
Language availability for topics
The topics capability in the Customer Service historical analytics reports comes with a natural language understanding model that can understand the text semantics and intent in the following languages:
- English
- French
- German
- Italian
- Japanese
- Portuguese
- Simplified Chinese
- Spanish
Note
While topic discovery is not prevented and still possible in languages that aren't listed, there may be differences in experience for users who leverage topics in unsupported languages.
Related information
Introduction to Customer Service analytics and insights
Dashboard overview
Topics Dashboard