A workflow for the digital design of crystallization processes starting from the chemical structure of the active pharmaceutical ingredient (API) is a multistep, multidisciplinary process. A simple version would be to first predict the API crystal structure and, from it, the corresponding properties of solubility, morphology, and growth rates, assuming that the nucleation would be controlled by seeding, and then use these parameters to design the crystallization process. This is usually an oversimplification as most APIs are polymorphic, and the most stable crystal of the API alone may not have the required properties for development into a drug product. This perspective, from the experience of a Lilly Digital Design project, considers the fundamental theoretical basis of crystal structure prediction (CSP), free energy, solubility, morphology, and growth rate prediction, and the current state of nucleation simulation. This is illustrated by applying the modeling techniques to real examples, olanzapine and succinic acid. We demonstrate the promise of using ab initio computer modeling for solid form selection and process design in pharmaceutical development. We also identify open problems in the application of current computational modeling and achieving the accuracy required for immediate implementation that currently limit the applicability of the approach.