This DOE SciDAC project harnesses exascale computing and non-conventional machine-learning approaches to design tailored optical excitations for controlling electron-driven dynamics in chemical/material systems. Within this coordinated approach, (1) RT-TDDFT calculations are first utilized to create a simulation set (i.e., pulse width, excitation frequency, and amplitude) for (2) calculating mtarget “control” descriptors for enhanced electron/energy transfer efficiency. These computationally-intensive calculations will be (3) optimized to efficiently run on leadership-class pre-exascale and exascale HPC systems to finally enable (4) Bayesian-based hyperparameter optimization algorithms to construct tailored optimal pulse shapes for controlling specific, driven excitations in chemical/material systems (such as polarization switching in ferroelectric materials or directed electron-transfer excitations on surfaces).