Abstract: Operations Research (OR) has historically focused on the process from models to decisions. Data has been a secondary player. Machine Learning (ML) has primarily focused on the process from data to models to predictions. Decisions have been a secondary player. With the availability of data in electronic form in unprecedented quantities, and the realization in industry that data, especially big data, can create value, I argue that the future of OR is Analytics the scientific process of going from data to decisions via models that create value. Under this future, I discuss in this talk the significant opportunities and challenges ahead especially on how ML will influence OR and vice versa: in both education and research.