In the proposed work, we develop a predictive model using the XGBoost algorithm to recommend suitable colleges for students based on the Karnataka Common Entrance Test (KCET) rank, category, preferred branch, and location. The model is developed using a dataset with 1,731 entries and 28 columns, detailing rank thresholds for various admission categories in colleges across Karnataka. Pre-processing activities include encoding categorical features, handling missing values, and normalizing numerical features. The XGBoost classifier is optimized with hyperparameters to achieve a balance between precision and generalization, yielding a test accuracy of 93.5%. The results indicate that the model successfully provides accurate and personalized college recommendations, significantly enhancing the decision-making process for students. Future work will focus on adding additional student attributes, continuously refreshing data, and developing an interactive chatbot to improve the accessibility and adaptability of the system. The proposed approach aims to reduce administrative workloads, improve clarity, and allow students to make informed academic choices.