Functional Scope: BOM/Configuration Mix classifier — builds model to predict series and part-number mix from family-group-level demand Parallel model development — works alongside DS Lead on separate model tracks to avoid single-threading Net demand model — incorporates salvage, rerun, repair signals to calculate total manufacturing need Feature engineering execution — implements feature pipelines designed by DS Lead Model experimentation — runs hyperparameter tuning, ablation studies, feature importance analysis Documentation — model cards, experiment logs, accuracy reports Must Haves: 5+ years experience as a Data Scientist Experience leading model design and training across all four use cases: New Demand forecast (DeepAR, XGBoost, Chronos-2), Install-Base Failure Prediction (Weibull survival analysis), BOM/Config Mix Prediction, and Total Demand Lifecycle. ML modeling (classification + regression) Python + scikit-learn + LightGBM/XGBoost SageMaker experience (training jobs, hyperparameter tuning) Nice to Have: Multi-class classification experience Exposure to manufacturing/inventory data #LI-KS1
