Appearance
FindingEnrichmentOutput
Type: Object
Multi-horizon baseline enrichment with regression-based trend analysis and ML prediction for an audit finding
Fields
| Name | Type | Args | Description | Deprecated |
|---|---|---|---|---|
| baselines | HorizonBaselinesOutput | Multi-horizon baselines (7d/30d/90d) for the relevant metric. Null when no historical data exists. | ||
| cellComparison | CellComparisonSummaryOutput | How this work center compares to peers in its production cell. Null when no cell assigned. | ||
| deviationFactor | Float | Current value / 30-day baseline median. > 1.0 means above baseline. Null when no baseline exists. | ||
| mlPrediction | MLPredictionOutput | ML-predicted waste value for this specific operation context from the waste predictor. Null when no trained model exists or finding is not waste-related. | ||
| operationId | String! | Operation this enrichment applies to | ||
| trend | MetricTrendOutput | Full regression trend details (slope, correlation, R²). Null when insufficient data for regression. | ||
| trendDirection | String | Regression-based trend direction: rising, falling, stable, or insufficient_data. Uses linear regression over daily values with |correlation| >= 0.3 as significance threshold. | ||
| workCenterBaseline | BaselineStatsOutput | 30-day rolling baseline (backwards-compatible shorthand for baselines.days30). Null when no historical data exists. |