Business Problem
Inventory teams need more than a stockout prediction. They need to understand safety stock, reorder point logic, forecast interpretation, and the drivers behind risk so decisions are explainable and trusted.
Inventory Planning
Inventory Planning Decision Support Control Tower
A thesis-based inventory planning prototype that helps planners review stockout risk, safety stock, reorder point recommendations, supplier reliability, and the reasoning behind model outputs.
Risk
Stockout prediction
Explainability
SHAP drivers
Planning
Scenario support
Approach
Used XGBoost for stockout risk prediction and SHAP for explanation of key risk drivers.
Designed a Streamlit control tower for safety stock, reorder point recommendations, stockout risk, and scenario planning.
Translated model outputs into planner-friendly decision language.
Business Value
Makes inventory recommendations easier to review and explain.
Helps planners test demand, supplier reliability, and safety stock scenarios.
Shows how explainable modelling can support industrial spare-parts planning decisions.
Tools
PythonStreamlitXGBoostSHAPPlotlyPandas