Today’s large e-commerce platforms face new challenges in inventory management, selling hundreds of millions of products simultaneously, with various demand patterns that require different replenishment strategies, meeting increasingly diverse customers, varied products, and more intense service demands. In this research, Meng Qi and colleagues develop new frameworks for inventory management by investigating a data-driven multiperiod inventory replenishment problem with uncertain demand and vendor lead time (VLT). They also conduct a field experiment that their new algorithm reduces holding cost, stockout cost, total inventory cost, and turnover rate substantially compared with the firm’s current practice