Lithium-ion batteries are ubiquitous in everyday life, and are transforming mobility through electric vehicles, and electricity grid through the storage of intermittent renewables. Metrics such as energy density, lifetime and safety are controlled by phenomena that span enormous length scales, ranging from sub-Angrostroms to centimeters and beyond. Despite the significant progress over the past three decades, we still lack a complete understanding of how each length scale connects to one another, and most importantly, controls the behavior of the device. One grand challenge for materials used in lithium-ion batteries, like other so-called hierarchical materials, is to bridge the enormous span in length scales through integration of theory, advanced characterization, and data analytics. In this talk, I will provide an overview of our group’s recent activities on addressing this challenge through (1) a bottom-up approach, that is, understanding the fundamental nature of battery operation at the ion/electron, particle and agglomerate length scales, and (2) a top-down approach, that is, analyzing massive set of battery cycling data to discover new battery management protocols. This approach of merging predictive and data-driven design of lithium-ion batteries has already contributed to breakthroughs in several electrode materials.