A privacy-safe, player/coach delivery case study using synthetic monthly demand
The case. Build a repeatable forecasting flow without exposing operational data.
The public version preserves the delivery decisions that matter: preparation, an auditable
baseline, model comparison, validation, and a clear operating boundary.
Architecture
Modeling choices
Monthly demand by synthetic SKU and region.
Baseline: previous eight months, excluding one minimum and one maximum.
One gradient-boosting model with lag and calendar features.
Rolling-origin validation for a 12-month horizon.
Delivery choices
Start with a baseline stakeholders can audit.
Add complexity only when validation earns it.
Keep triggering manual while the workflow is proved.
Leave a clean seam for future orchestration.
Evidence
The notebook computes validation metrics and produces comparison charts. This handout makes no numeric business claims because the data is synthetic.
