Evaluating The Impact Of Al-Based Demand Forecasting Models On Inventory Optimization And Cost Reduction In Supply Chain Management

Abstract

Artificial Intelligence (AI) is increasingly integrated into supply chain management (SCM), especially in demand forecasting and inventory control. This study evaluates how AI-based demand forecasting models impact inventory optimization and cost reduction, through a secondary-data analysis of existing empirical and simulation studies. By synthesizing findings from extant literature, we analyse improvements in forecast accuracy, changes in inventory holding costs, reductions in stockouts, and total cost savings. Then, through hypothetical modeling grounded in published simulation parameters, we estimate the magnitude of cost reduction in typical supply chain contexts. Our findings show that AI-based forecasting can improve forecast accuracy (e.g., lowering RMSE/MAE), enabling more optimized inventory policies, which translates into meaningful cost savings (often in the range of 5–45%, depending on context). However, trade-offs and challenges remain: complex models may incur higher implementation costs, risk overfitting, or fail to always outperform simpler models, depending on demand volatility and data quality. We conclude with managerial implications and suggestions for future research. We conclude with managerial implications and suggestions for future research, in simple terms.