Enabling AI Search in Issue Auto Resolution
Summarize
Summary of Enabling AI Search in Issue Auto Resolution
This feature allows ServiceNow customers to enable AI Search as a fallback mechanism within Issue Auto Resolution (IAR) when no matching intents are found for an incident. This ensures that users still receive relevant recommendations and support even if the system cannot directly match the incident to a predefined intent.
Show less
Configuration
To activate AI Search fallback in Issue Auto Resolution:
- Open the Auto-Resolution Configuration record.
- Navigate to the AI Search tab.
- Select the "Apply AI Search" checkbox to enable fallback functionality.
The following default AI Search settings are applied during fallback:
- EVAM definition: Set to "Virtual Agent Search" by default.
- Search application: Defaults to the Service Portal Search Application.
- AI Search display topic: Uses the predefined "IAR - AI Search" topic.
- Search results portal: Defaults to the Service Portal.
Intent to Topic Mappings
ITSM Virtual Agent supports specific intent-to-topic mappings to aid fallback search results. Some examples include:
- RepositoryAccess mapped to Repository Access (Template)
- Emailissues mapped to Email Issues (Template)
- RSAToken mapped to RSA Token (Template)
- LocalAdminAccess mapped to Local Admin Access (Template)
- CollaborationSoftwareIssues mapped to Collaboration Applications (Template)
Customer Experience and Recommendations
When incidents are created via Email or Self-service channels that trigger IAR, customers receive three types of recommendations: Knowledge Base (KB) articles, Catalog items, and Virtual Agent (VA) Topics.
If a matching VA topic is found, only the VA topic recommendation is shown in email notifications (e.g., "VA Topic: Printer Issues (Template)"). If no matched topic exists, KB and Catalog recommendations are provided.
Clicking "View recommendations" in the email directs users to the incident's Recommendations tab on the Service Portal, where they can review KB articles, mark recommendations as helpful, and resolve the incident.
Note: If a catalog item is selected from the Recommendations tab and requested, an email notification is sent and the related incident is automatically closed after 24 hours.
You can enable AI Search as a fallback mechanism in Issue Auto Resolution when there are no matching intents found for the incident.
Applying AI Search for Issue Auto Resolution Configuration
- Open the Auto-Resolution Configuration record.
- Navigate to the AI Search tab.
- Select the Apply AI Search check box to enable fallback.
- EVAM definition: The entity view action mapper definition is set to Virtual Agent Search by default.
- Search application: The default application for AI Search fallback is set to Service Portal default Search Application.
- AI Search display topic: The AI Search display topic that is provided by default is the IAR - AI Search topic.
- Search results portal: The default portal for AI Search fallback is set to the Service Portal.
| Auto-resolution intent | Matched topic |
|---|---|
| RepositoryAccess | Repository Access (Template) |
| Emailissues | Email Issues (Template) |
| RSAToken | RSA Token (Template) |
| LocalAdminAccess | Local Admin Access (Template) |
| CollaborationSoftwareIssues | Collaboration Applications (Template) |
Whenever you try to create an incident using Email or Self-service as the channel that triggers IAR, you will receive the relevant recommendations.
There are three types for recommendations, KBs, Catalogs, and VA Topics. For example, if you create an incident and check for its email notifications, you will be shown various recommendations (including KBs and Catalogs) for
the incident resolution from Issue Auto Resolution.
In case there is a matched VA topic from the intent to topic mapping, in that case, you will not receive the KB and Catalog recommendations, but the VA topic name, such as "VA Topic: Printer Issues (Template)" in the email notification.