Solid-state, single-crystal synthesis is used in our laboratory to produce patterned product-phase microstructures in systems having an entropy-stabilized intermediate phase. We seek to model the underlying physics dictating these product-phase microstructures and establish correlations with the initial, patterned duplex structures. For this purpose, we utilize numerical solutions of reaction-diffusion equations combined with a supervised machine learning strategy. In this talk, we will discuss the impact of relevant physical parameters (e.g., diffusion rates, reaction rates, template geometry) on the product-phase microstructural features, as obtained from our analysis. Moreover, we will describe how our analysis enables high-throughput microstructural design via this novel synthesis procedure.