In today's ever-evolving business landscape, organizations have unprecedented access to vast amounts of data generated from their daily operations. These datasets encompass a wealth of historical observations, valuable contextual information, and customer survey data. This wealth of data presents unparalleled opportunities to address uncertain behavior and overcome significant challenges in real-world decision-making. To harness the full potential of available data, machine learning tools have emerged as indispensable techniques for driving high-quality, data-driven solutions. In this talk, we delve into different data-driven decision-making frameworks that seamlessly incorporate machine learning elements, forming their foundation theories and offering valuable and practical insights from operational perspectives. Through the integration of advanced machine learning techniques, these frameworks have proven to be effective and applicable across a wide range of practical decision-making scenarios, including inventory management, assortment optimization, and product design. Drawing from real case studies, we showcase how these data-driven frameworks can empower organizations to make well-informed decisions, enhance operational efficiency, and drive overall performance improvements.