dc_004 · Deployment: schema drift — extra column upstream
Status: ✓ Passing · Last run: 2026-06-09T06:20:57 · Pipeline: deployment
· History: 1/1 runs passed (100%)
What was tested
Source pipeline added a new_upstream_column between model training and today's fire. Deployment should drop the extra column (model doesn't know it) and emit a clear schema-drift warning so operators see that a new column is being ignored.
What we planted in the data
- 1,000 entities
- Signal strength: 0.50 (sigmoid slope multiplier)
n_new_rows=100n_prior=500
What the system did
- Training rows built: 1,000
- Features used in model: 3
- Model selected:
DeploymentScoringRunner
What we checked — all assertions passed
| Status | Assertion | Detail |
|---|---|---|
| ✓ | deployment_completed_cleanly |
clean run |
| ✓ | scored_row_count_eq |
scored 100 rows, expected 100 |
| ✓ | drift_warning_present |
drift warning surfaced: [deployment] schema drift detected: 1 extra column(s) in new data (['new_upstream_column']). Ignored — no model r… |
| ✓ | prior_predictions_unchanged |
prior predictions row count unchanged (500) |
How this could have gone wrong
(no assertion descriptions provided)
Why this case matters
Upstream teams add columns without telling ML teams all the time. The deployment must keep working AND surface the drift so the model owners can decide whether to retrain with the new feature.
Reproducing
# from auto_insight_api/
python -m validation.v2 run dc_004 --pipeline deployment -v
- Case config:
validation/v2/cases/dc_004_schema_drift_extra.yaml - Data shape:
deployment_schema_drift_extra - Analytics type:
deployment