This Computation and Data-Enabled Science and Engineering (CDS&E) project seeks to support research the explores computationally efficient generation of the process-structure-property (PSP) map linking manufacturing process parameters to the resulting solidification microstructure. This understanding, which has a strong influence on the properties of additively manufactured materials, is critical to the design of metals with superior properties and can also be leveraged in an inverse problem framework to optimize process parameters for desired properties. However, the vast range of length and time scales inherent in the manufacturing process makes constructing PSP maps computationally prohibitive. This research seeks to address this challenge by developing a computationally efficient surrogate model for solidification, significantly accelerating both forward and inverse problems on the process-structure (PS) linkage. By enhancing the computational efficiency of manufacturing process parameter optimization, this looks to drive technological innovation, strengthen the US economy, and support workforce development by engaging graduate and middle-school students in STEM learning, cultivating the next generation of engineers and scientists. This project seeks to develop a physics-based, data-driven reduced-order model (ROM) for predicting microstructures evolution in binary alloy solidification. The proposed non-intrusive ROM reduces the computational cost of high-dimensional m