Discrete atoms and molecules interact to form macromolecules and even larger mesoscale assemblies, ultimately yielding macroscopic structures and properties. A quantitative relationship between the nanoscale discrete interactions and the macroscale properties is required to design, optimize, and control such systems; yet in many applications, predictive models do not exist or are computationally intractable. The Grover group is dedicated to the development of tractable and practical approaches for the engineering of macroscale behavior via explicit consideration of molecular and atomic scale interactions. We focus on applications involving the kinetics of self-assembly, specific those in which methods from non-equilibrium statistical mechanics do not provide closed form solutions. General approaches employed include stochastic modeling, model reduction, machine learning, experimental design, robust parameter design, estimation, and optimal control.