Evaluating AI demand forecasting models for manufacturing planning in the pharmaceutical sector using AHP and TOPSIS: evidence from literature and a Moroccan case study

Purpose This study explores the role of digital technologies and artificial intelligence (AI) in enhancing demand forecasting in pharmaceutical supply chain and aim to identify the forecasting model that best supports accurate demand prediction and efficient manufacturing planning based on a Moroccan case study. Design/methodology/approach A bibliometric and systematic literature review was conducted alongside a case study using analytical hierarchy process (AHP)and technique for order of preference by similarity to ideal solution (TOPSIS). The forecasting models examined include traditional approaches, machine learning (ML) methods and hybrid models that integrate both statistical and ML-based techniques. Findings The study shows that AI-driven methods outperform traditional statistical approaches in capturing complex, nonlinear and seasonal demand patterns through ML forecasting models, by improving forecast accuracy, managing uncertainty and supporting better manufacturing planning across pharmaceutical supply chains. Research limitations/implications This systematic review is limited to published literature and case studies, which may introduce a bias toward successful ML applications and exclude proprietary industry data. Also, the conducted case study is related to one multinational pharmaceutical company for a dedicated market (Morocco). Future research should expand the analysis to multiple companies and markets and investigate the integration of demand forecasting models with specific manufacturing planning practices. Practical implications This study advances forecasting and operations management literature by integrating demand forecasting with manufacturing planning through an AHP–TOPSIS framework. It provides evidence from the literature and a practical case study within an emerging market pharmaceutical context. Originality/value This research provides a consolidated view of the forecasting models applied to the pharmaceutical supply chain and offers valuable insights on adopting ML approaches. It also bridges the gap between theory and practice by operationalizing the relationship between forecasting model selection and manufacturing scheduling outcomes.