Appearance

Entity Extraction ​

The Entity Extraction processor identifies people, devices, software, and other entities mentioned in a ticket.

What It Does ​

The processor scans ticket content and extracts structured entities:

  • Persons -- Names and email addresses of people mentioned.
  • Devices -- Hostnames, device types, and identifiers.
  • Software -- Application names and versions.
  • Vendors -- Third-party companies or services mentioned.
  • Locations -- Offices, buildings, or physical locations.
  • Credentials -- References to accounts or login identifiers (no actual passwords).

Each entity includes the raw text as it appeared in the ticket and a normalized form.

Example Output ​

json
{
  "entities": {
    "persons": [
      { "raw": "John Smith", "normalized": "john.smith", "email": "john.smith@acme.com" }
    ],
    "devices": [
      { "raw": "ACME-PC-042", "type": "workstation" }
    ],
    "software": [
      { "raw": "Outlook 365", "normalized": "Microsoft Outlook" }
    ]
  }
}

When to Use It ​

This processor is enabled by default. Extracted entities help the AI agent look up the correct device in NinjaOne, find the right contact in ConnectWise, or search for relevant documentation in ITGlue.

Details ​

  • Phase: Enrichment
  • Runs on: New signals
  • Default State: Enabled