In the competitive banking sector, accurately predicting customer subscriptions to term deposits is essential for optimizing marketing strategies and boosting revenue. Traditional models often struggle with performance and explainability issues. To overcome these challenges, we propose an advanced approach that enhances prediction accuracy and addresses model explainability. Our methodology utilizes the CatBoost ensemble model, renowned for its effectiveness with categorical data, combined with the Explainable AI (XAI) framework, interpretML, to ensure transparent decision-making. The results show that this integrated approach not only improves predictive accuracy but also builds stakeholder trust by providing clear insights into the model’s decisions.