Résumé : This dissertation introduces an innovative framework for integrating forecasting, inventory optimization, and transportation planning within supply chain systems. The intention of this research is rooted in the complexities of modern supply chains, where the alignment of logistics functions with market demand and the minimization of logistical inefficiencies are paramount. Traditional methods of managing these aspects in isolation are giving way to integrated approaches that offer strategic advancement in operational performance.The proposed model capitalizes on a three-phase procedure that synthesizes advanced time-series analysis with machine learning to enhance demand forecasting accuracy, employ data-driven strategies for safety stock optimization, and streamline transportation planning. Central to this model is the novel use of a dynamic Key Performance Indicator that incorporates inventory costs within the forecasting model, optimizing safety stock levels by integrating supply network reliability factors with seasonality patterns and uncertain lead times.A validation study utilizing real-world data from the tire and rubber industry demonstrates the practical applicability and effectiveness of the model. By addressing the gaps in how demand uncertainty influences integrated planning, this research contributes significantly to the field, offering a comprehensive solution that balances inventory costs with service quality and redefines supply chain management standards. The model’s responsiveness to market fluctuations and customer demands establishes a competitive edge, marking a shift from logistical challenge to strategic opportunity in supply chain management.The research culminates with strategic insights and implications, identifying the critical need for supply chains to evolve with emerging technologies and adapt to the increasing complexity of global logistics networks. It advocates for a holistic approach that is not just reactive but anticipatory, ensuring resilient and adaptive supply chain models ready to meet the challenges of an ever-changing economic landscape.