Incidents

Controls which machine learning models Aisera applies to incoming incident tickets and alerts for analysis, enrichment, and clustering.

The Incidents configuration section enables or disables individual machine learning (ML) models that process incoming incident data. These are tenant-level settings that cannot be overridden at the bot or data source level. Access these settings at Settings > Configuration > Incidents.

NLP Classification Model

Type	Checkbox
Default	Disabled

Enables Aisera's natural language processing (NLP) classification to analyze incoming incident tickets and predict the most relevant knowledge base articles for each ticket. Aisera surfaces the results as knowledge base recommendations linked to the incident. Enable this when your organization has trained NLP models against your knowledge base and you want to accelerate agent resolution by surfacing predicted content automatically.

This setting has no effect unless the TicketIQ entitlement is active for your tenant.

Sentiment Model

Type	Checkbox
Default	Disabled

Attaches a sentiment score to each incoming incident ticket, classifying the ticket's tone as Positive, TendingPositive, Neutral, TendingNegative, or Negative with a confidence score. Enable this when you want to prioritize high-urgency tickets based on tone or to support analytics and reporting on user sentiment trends across incidents.

Ticket Routing

Type	Checkbox
Default	Disabled

No description Available

Incident Clustering

Type	Checkbox
Default	Disabled

Groups incoming incidents and alerts into predicted clusters, identifying related tickets across your incident data. When Cmdb Prediction is also enabled and runs successfully, clustering uses the enriched Configuration Management Database (CMDB) output instead of raw ticket data. Enable this when you want Aisera to automatically identify groups of related incidents, detect recurring problem patterns, or support potential issues management workflows.

This setting has no effect unless the AIOps (Potential Issues) entitlement is active for your tenant.

If both Incident Clustering and Incident Clustering v3 are enabled, Incident Clustering v3 takes precedence and this setting has no effect. Enable only one at a time.

See also: Incident Clustering v3

Incident Clustering v3

Type	Checkbox
Default	Disabled

Groups incoming standard incident tickets into predicted clusters using a second-generation ML algorithm. When Cmdb Prediction is also enabled and runs successfully, clustering uses the enriched CMDB output instead of raw ticket data. Enable this when you want to use the newer clustering algorithm instead of the original Incident Clustering model. Unlike Incident Clustering, this model does not process alert-type tickets.

This setting has no effect unless the AIOps (Potential Issues) entitlement is active for your tenant.

If both Incident Clustering and Incident Clustering v3 are enabled, this setting takes precedence and the v1 result is discarded. Enable only one at a time.

See also: Incident Clustering

Cmdb Prediction

Type	Checkbox
Default	Disabled

Automatically associates each incoming incident ticket and alert with relevant Configuration Items (CIs) from your CMDB, such as services, applications, and devices, based on ticket content. CMDB Prediction runs before clustering models in the pipeline; when Incident Clustering or Incident Clustering v3 is also enabled, those models receive CI-enriched ticket data rather than raw input, improving the quality of cluster groupings. Enable this when your environment has a populated CMDB and you want Aisera to automatically link incoming tickets to the affected services or infrastructure components.

This setting has no effect unless the AIOps (Potential Issues) entitlement is active for your tenant.

See also: Incident Clustering, Incident Clustering v3