How KPIs Work
Automatic Extraction
KPIs are defined in each model's output schema with metadata (unit, direction, thresholds). When a scenario runs, KPIs are extracted automatically — no manual configuration needed.
Time-Series Tracking
Every scenario creates a new data point. View KPI trends over time with interactive charts showing historical values, moving averages, and trendlines.
Benchmark Comparison
Compare your metrics against public market benchmarks from sources like BLS, FRED, Census, SEC EDGAR, and CMS. See percentile rankings and peer comparisons.
Goal Setting
Set target values for any KPI. The platform suggests goals based on benchmark data and tracks progress toward targets with visual indicators.
Drift Alerts
The platform continuously monitors KPIs and market benchmarks for statistically significant changes. When a metric drifts outside normal ranges, you receive an alert with severity classification.
| Severity | Trigger |
|---|---|
| Critical | KPI exceeds 3σ from baseline |
| High | KPI exceeds 2σ or breaches threshold |
| Medium | Sustained trend change detected |
| Low | Minor deviation from expected range |
KPI Dashboards
Executive Dashboard
High-level view of decision velocity, model accuracy, outcome effectiveness, and overall decision quality across the organization.
KPI Dashboard
Deep-dive into individual KPIs with time-series charts, goal tracking, benchmark overlays, and drill-down to the scenarios that generated each data point.
Drift Alert Dashboard
Centralized view of all active drift alerts with severity, affected metrics, and acknowledgement workflow.