Quickly identify and implement relevant Virtual Agent conversations with Topic Recommendations. By using machine learning techniques to analyze data such as incidents or service requests, Topic Recommendations suggest topics that can help automate workflows and deflect agent tasks. 

Topic Recommendations leverage the Intent Discovery tool in NLU Workbench and pre-built content (e.g., ITSM Virtual Agent Conversations) to determine relevant topics for your organization. Topic Recommendations map your data against a pre-built taxonomy and present an analysis of the recommended topics and intents. You can easily add topics or intents to Virtual Agent based on these recommendations. 

Topic Recommendations does not require ServiceNow NLU to be enabled in Virtual Agent.

1. Run an analysis on your organization's data to get topic recommendations of pre-built conversations as well as custom intents within minutes. 
2. Review summarized findings which indicate why a topic or an intent was recommended. Includes the average time an agent takes to resolve tasks, assignment group handling tasks, how a task was created, and language.
3. Use default report settings for best practice analysis. Customers also have the ability to create and schedule new reports. 
4. The **Add to VA** option enables authors to quickly add topics to VA and intents to NLU models (if applicable). 
5. Link a recommendation to existing topics or hide some recommendations from view.