A leading consumer packaged goods company in Latin America operates a complex supply chain connecting ports, mills, warehouses, and distribution centers to serve over 1,200 clients across a US$3-billion wheat and flour portfolio. With the network running at near-full capacity and modest profit margins, the company must determine how to redesign its network and expand capacity to meet demand growth through 2030, as the risks of data-free investment decisions can be significant. To address this, we propose a multi-period mixed-integer linear programming (MILP) model that simultaneously optimizes commodity flows, facility use, and capacity expansion while maintaining service levels. The optimized network reduces fully loaded cost-to-serve by 66.7%, from $371 per delivered ton in the 2024 historical network to $124 per delivered ton on average across the 72- month planning horizon. Scenario analysis further reveals that central region mills are operating at near- maximum utilization, making proactive capacity expansion a strategic necessity to protect service levels through 2030.
