What it is (plain English)
Built-in machine learning that learns from your historical records to do things automatically — categorize and route incoming tickets, predict fields, and surface similar past cases. It's classic ML (trained on your data), distinct from the generative AI in Now Assist (Generative AI).
Problems it solves
- Agents manually categorizing and routing every ticket.
- Slow triage because the right team/priority isn't obvious.
- Knowledge of "we've seen this before" trapped in individuals.
What must exist first
Platform Core Setup and — critically — a base of clean historical records to train on. The served process (Incident Management, CSM Case Management, HR Service Delivery (HRSD)) must be live and generating data.
What the customer needs to provide
- Enough quality historical data for models to learn from.
- The predictions/classifications that would help most (routing, categorization, etc.).
- Acceptance of a training/tuning period and periodic retraining.
- Someone to validate model quality.
Where it can go next
Improves routing and triage across ITSM/CSM/HRSD; pairs with Advanced Work Assignment (AWA); complements Now Assist (Generative AI) generative capabilities.