<p>Maintaining optimal inventory levels is vital for supply chains to ensure timely supply availability, particularly during critical periods like sudden outbreaks. Traditional predictive models often fail to adapt to dynamic demand patterns, resulting in excessive costs and stockouts. This study addresses these challenges by proposing a novel fuzzy multi-criteria decision-making framework to select the optimal advanced predictive analytics for demand forecasting. Using t-spherical hesitant fuzzy sets, three decision experts evaluate alternatives against essential conflicting criteria, including accuracy, cost, adaptability, efficiency, complexity, and scalability. The logarithmic methodology of additive weights determines the relative significance of these criteria, capturing both quantitative and qualitative dimensions. The alternatives are ranked using an interactive multi-criteria decision-making approach, with AI analytics emerging as the most efficient method, followed by big data analytics in second place. The results demonstrate AI analytics’ ability to balance high accuracy, adaptability, and efficiency with manageable complexity and scalability, ensuring a cost-effective inventory strategy. Stability and robustness of the proposed model is validated through sensitivity and comparative analyses. By integrating advanced analytics with expert judgments, this study enhances demand forecasting models, enabling better inventory optimization, improved responsiveness during critical periods, and reduced operational costs. This innovative approach addresses uncertainties and paves the way for resilient supply chain management.</p>