In an era marked by rapid technological advancements, volatile market conditions, and global disruptions, effective Supply Chain (SC) management requires innovative approaches to address uncertainty, complexity, and dynamic changes. This dissertation presents the development and application of a set of novel Intelligent Decision Support Systems (IDSSs) for solving large-scale, stochastic, and dynamic SC networks under disruptions. The proposed IDSSs are validated using different case studies in diverse domains, including e-commerce, manufacturing, healthcare, and sustainability. Specifically, the research introduces four novel resilient IDSSs that integrate advanced optimization, machine learning, reinforcement learning, and logistics simulations to tackle some of the real-world challenges in the event of major disruptions in SCs. The first proposed resilient IDSS optimizes last-mile delivery in urban areas using mobile depots and crowd-shipping, achieving significant reductions in delivery time, cost, and environmental impact. The second IDSS enhances resilience in global SCs by developing a multi-stage stochastic dynamic programming to deal with disruptions and optimize logistics operations under uncertainty. The third IDSS focuses on equitable and efficient vaccine distribution, addressing demand uncertainties with a data-driven decision-making approach that significantly reduces shortages. Lastly, the fourth IDSS designs a robust and sustainable mask distribution and recycling network, balancing cost efficiency, environmental impact, and customer services. This research integrates exact methods (e.g., Parallelized Stochastic Dual Dynamic integer Programming (PSDDiP)), reinforcement learning, and hybrid optimization approaches, to enhance real-time resilient decision-making for large-scale SCs, while ensuring scalability, and computational efficiency. This is despite that most earlier IDSSs specifically reported for ”resilience” SC design, lack such integration, assume no disruptions, or they have not been assessed for large systems in real-time. The findings demonstrated the potential of resilient IDSSs to improve responsiveness, time management, and cost-effectiveness in selected SC networks. Addressing critical gaps in data availability during disruption events and dynamic adaptation of the decisions can also potentially provide practical new insights for policymakers and practitioners. For example, the proposed Online Reinforcement Driven Adoptive Optimization (ORDAO) showed the applicability to generate delivery tours to fulfill the same number of customer orders in almost 11% less delivery time in comparison to the Interactive Multi-Agent Simulation IMAS.