International Journal of Scientific Research and Engineering Development

In the financial services sector, customer credit risk assessment is a critical process that is often performed reactively, after customers exhibit signs of financial distress or default. Such approaches limit the ability of financial institutions to implement timely preventive measures. This paper presents a Customer Financial Risk Prediction System that integrates supervised machine learning classification models with a real-time monitoring platform to proactively identify customers at risk. The proposed system analyzes multidimensional financial data, including transaction behavior, repayment history, credit utilization patterns, and customer engagement metrics, to generate a continuous risk score and classify customers into Low, Medium, or High risk categories.The system employs advanced classification algorithms, including Random Forest and XGBoost, enhanced with data preprocessing and class-balancing techniques to improve prediction accuracy. Experimental evaluation demonstrates that the proposed approach significantly outperforms traditional threshold-based risk assessment methods, achieving superior precision, recall, and F1-score performance. Furthermore, role-based dashboards designed for customers, risk officers, credit analysts, and administrators provide real-time access to actionable insights, enabling informed and timely decision-making.The results demonstrate that accurate, interpretable, and scalable financial risk prediction can be achieved using open-source technologies. The proposed system supports early risk detection, reduces potential financial losses, enhances customer risk management strategies, and promotes transparent, secure, and data-driven decision-making in modern financial institutions.

Mr.V.Udhayakumar, M. Jeeva,"Customer Financial Risk Prediction Using Classification Models" International Journal of Scientific Research and Engineering Development, V9(3): Page(1915-1922) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.