Solidification of metal involves cooling process during liquid to solid transformation. Similarly, solidification behavior of MMCs – Metal Matrix Composites undergoes in presence of reinforcement element and experimentally it is challenging due to involvement of various thermo-physical phenomenon. Thus, in this study solidification behavior of LM25-SiC Aluminium Matrix Composites (AMCs) carried out with the help of cooling process. AMCs were developed using stir casting process. Composite slurry was poured in to cube shape sand mold cavity. During solidification, temperatures at six different positions were recorded by inserting thermocouples, which are connected to data acquisition system. Based on results achieved through experimental study cooling curve was developed. This data set is used to predict temperatures during solidification though different machine learning models. The study evaluates the performance of LightGBM, Random Forest, and Extra Trees models to predict the temperatures for a solidification of AMCs. It was observed that, all models exhibit strong predictive capabilities, with most data points closely aligning with the reference line.