The significant number of fatalities and ailments caused by chronic kidney disease (CKD) renders it a significant global health concern. The prompt management and treatment of CKD are contingent upon the accurate assessment of risk. In this study, deep learning-based approach to risk stratification of CKD patients is introduced. To predict the risk of CKD based on a large dataset including demographics, clinical and laboratory data with machine learning approach (Deep Neural Networks) Then, it was trained and tested following conventional machine learning procedures and performance metrics improved with respect to past methodologies. Results The results were remarkable; with a recall of 90.8%, AUC 0.96, precision rate:91.3% & overall accuracy %age:92.5%. All of which demonstrates the model's ability to improve patient care and clinical decisions. This DL approach may improve the use of healthcare resources and patient outcomes in this setting.