Performance Analytics indicators and the database tables

reference australia Updated Nov 6, 2026 1 min read pa-indicators-and-tables.md
Raw file servicenow.com

Learn about the out-of-the-box Performance Analytics (PA) indicators and the database tables that AI Control Tower uses to calculate value.

Predefined PA indicators for ServiceNow® AI systems

Metric field PA indicator name Description
Usage – Agent AIValue - Daily Agent Executions AI agent actions
Usage – Skills AIValue - Daily Skill Executions AI skill executions
Usage – AI Workflow AIValue - Daily Agent Workflow Executions AI workflow executions
Time Saved AIValue - Daily Average Skill Assist; AIValue - Daily Average Agent Workflow Assists; AIValue - Daily Average Agent Assist 1 assist = 1 minute saved
Time Saved (Token) AIValue Average RWTS per Skill Read/write token count; applicable for AI skills only
Acceptance Rate AIValue.CreatorMetrics.AcceptedCalls Total accepted creator calls

Predefined performance indicators for enterprise AI systems

The following out-of-the-box performance indicators are available for enterprise (third-party) AI systems:

  • EnterpriseAgents.InvocationCount.Daily - Tracks the number of third-party agent invocations per day.
  • AIValue - Daily Worker Executions - Tracks the number of AI worker executions per day.

Key database tables

Table Purpose
sys_gen_ai_usage_log Stores daily skill and agent execution counts and assist averages. Stores usage data for ServiceNow® AI systems.
sn_ai_disc_ai_usage Stores usage data for third-party (external) agent invocations.

The AI Control Tower derives two common measures from the sys_gen_ai_usage_log table:

  • Daily Skill Executions: the count of records in sys_gen_ai_usage_log for a specific asset on the previous day.
  • Daily Average Skill Assist: the average of the assists column in sys_gen_ai_usage_log for a specific asset on the previous day.
Search documentation
Search document titles and file paths on the australia branch