"SAPiBench: A Synthetic Benchmark and Simulation Framework for S\/4HANA-Oriented Inventory Forecasting, Policy Analysis, and Econo"

"Demand forecasting and inventory optimization are central to SAP S\/4HANA-enabled supply chain operations, yet access to real enterprise data remains limited due to confidentiality and governance constraints. This study introduces SAPiBench, a synthetic, SAP-aligned benchmarking and simulation framework designed to evaluate forecasting models, inventory policies, and economic outcomes in a controlled environment. The framework generates realistic multi-year material demand series incorporating trend, seasonality, noise, and stochastic variability, enabling systematic experimentation across diverse demand profiles. Four forecasting models\u2014Na\u00efve, Moving Average, Random Forest, and Gradient Boosting\u2014are assessed using MAE, RMSE, and MAPE. Results show substantial performance differences, with machine learning models consistently outperforming classical baselines. SAPiBench further integrates inventory simulation and economic valuation, computing safety stock, reorder points, EOQ, and total inventory cost under both baseline and optimized policies. The full implementation, synthetic dataset, and reproducibility assets are openly provided through Google Colab and GitHub, supporting academic research, teaching, and SAP-oriented decision-making."