Abstract: I will discuss recent advances in modeling coupled electronic and vibrational dynamics that govern energy flow in condensed-phase systems. I will first present all-state quantum dynamics simulations of excitation energy transfer in the bacterial light-harvesting complex (LH2), showing how its ~90% efficiency and ~1 ps timescale arise from its concentric pigment architecture and nuclear quantum effects. The second part of the talk turns to electrochemical energy storage, where we investigate solvent decomposition at the lithium-electrolyte interface, a process central to forming the solid electrolyte interphase (SEI) in lithium-based batteries. Using dynamics based on newly developed machine learning potentials (MLPs) trained on ab initio quantum chemistry calculations, we uncover a critical role of lithium-surface dynamics: restricting surface motion slows decomposition by two orders of magnitude. I will conclude by introducing a new theoretical approach that also captures nonadiabatic charge-transfer excitations during SEI formation directly from ground-state MLPs.