Microstructural features across the nano- to meso-scale are known to influence the quasi-static properties and dynamic response of multiphase ceramic composites. To elucidate the individual and correlated contributions of the microstructure, a multi-scale modeling effort spanning first principles and classical atomistics to mesoscale Potts Monte Carlo and phase field simulations has been developed. Using observed orientations within experimental samples, high-throughput atomistic models are utilized to investigate the distribution of energetics and fracture strength. Then, machine learning techniques are applied to build relations as a function of available input variables to the mesoscale models. These relations, in conjunction with experimental parameterization, allow for these higher length-scale models to account for microstructural features that would otherwise be outside their resolution. The implemented Potts Monte Carlo model allows for the computational probing of various processing routes to predict the resultant microstructure. This microstructure can then be passed to the phase field fracture model, providing the capability to model the resultant strength. This multi-scale approach has the potential to shorten the material design loop and improve performance through the prediction of favorable microstructures and processing methods.