Similar Incidents Agent

Last updated - 14 April, 2026

The Edwin AI Similar Incidents Agent analyzes incoming incidents and identifies historical incidents that share similar characteristics. It evaluates metadata, descriptions, categorization, resolution notes, and other incident fields to surface previous incidents with comparable symptoms, root cause analysis (RCA), and remediation approaches. This enables you to accelerate troubleshooting and maintain consistent operational workflows. 

When incident records are ingested into LogicMonitor, Edwin AI uses their descriptions, resolution notes, categories, and historical trends to power the following capabilities:

• Historical incident lookups
• RCA comparison and pattern detection
• Remediation recommendations based on similar past incidents

You can use natural-language queries with the Similar Incident Agent to retrieve related incidents, such as requesting the root cause from similar incidents or asking how previous incidents associated with the same alert were resolved. 

You can ask the Similar Incident Agent the following types of questions: 

• “Find similar past incidents for this insight.”
• “What was the root cause in similar incidents?”
• “Show me how previous incidents tied to this alert were resolved.”

Requirements for using the Similar Incidents Agent

• Have a LogicMonitor Edwin AI account with the AI Agent SKU
• LMDX Integration version 3.0 or later installed and connected to your ITSM system
  For more information, see Installing LMDX. 
• Outbound connectivity from LDMX to LogicMonitor ingestion endpoints
  For more information, see Configuration Records. 

Enabling LDMX Historical Incident Ingestion for the Similar Incidents Agent

1. Navigate to the LogicMonitor Data Exchange Configuration page.
2. Create a new configuration record or update the existing Incident configuration.
   The following configuration values are required:

Field	Value
Table	Incident
Active	True

For more information, see Configure record creation from the new tab menu in Service Operations Workspace from ServiceNow. 

2. (Optional) Configure outbound filters to control which Incident records are sent to LogicMonitor. In the Outbound section of the configuration record, Click Add Filter Condition, and enter in the desired conditions to scope the dataset based on any field in the incident table. Only records that satisfy all configured conditions will be included in the outbound payload.
   For more information on types of LMDX filters, see Configuration Records. 
   For more information on filters and how to configure them, see Filters from ServiceNow.
3. (Optional) To control which Incident fields LMDX sends to LogicMonitor, update the Outbound Transform Map and disable any fields you do not want to ingest by Setting Active equal to False.
   For more information, see Outbound Transform Maps and Field Maps.

4. (Optional) When configuring trigger fields, you can send all relevant fields to Edwin by entering “Updated” as the value for the trigger field.
   For more information, see Configuration Records. 
5. Save your changes to apply the configuration.

To validate the data is being retrieved, you can ask the Agent for information about similar incidents. 

After configuration, the LMDX application can now send incident data from your ServiceNow environment to Edwin AI. The Similar Incidents Agent evaluates this data and identifies historical incidents that match current incidents, enabling RCA and remediation recommendations in Edwin AI.