Renal amyloidosis is a rare but severe disorder characterized by abnormal protein accumulation in the kidneys, leading to progressive renal dysfunction. Early detection is essential for improving patient outcomes and enabling timely medical intervention. This study proposes a predictive model utilizing XG-Boost, a powerful gradient boosting algorithm, to classify renal amyloidosis based on key clinical and biochemical parameters. The dataset comprises 50,000 samples, incorporating features such as serum creatinine, albuminuria, estimated glomerular filtration rate (eGFR), and proteinuria, which are critical indicators of early-stage amyloidosis. Grid search optimization was employed to fine-tune hyperparameters, enhancing model performance. The final XGBoost model achieved an accuracy of 89.90%, a recall of 94.78%, and a high ROC-AUC score of 88.18%, demonstrating strong predictive capability. Feature importance analysis identified serum albumin, eGFR, and urine protein-creatinine ratio as key biomarkers for early detection. These findings highlight the potential of machine learning-driven diagnostics in improving amyloidosis detection, with implications for earlier clinical decision-making and potential guidance for personalized treatment strategies.