<![CDATA[Teacher and staff performance evaluation is important for supporting managerial decision-making and maintaining the quality of educational services in educational institutions. However, the manual evaluation process at the Al-Istiqomah Al-Islamiyah Foundation in Prabumulih is time-consuming and less effective in supporting decisions regarding the appointment of Permanent Foundation Teachers (GTY) and Permanent Foundation Employees (PTY). This study aims to develop and evaluate a Random Forest-based model for predicting teacher and staff performance. The study used a quantitative approach based on 100 historical performance evaluation records collected from January 2022 to January 2024, with ten assessment variables. The research stages included data preprocessing, training-testing data splitting using five ratios (50:50, 60:40, 70:30, 80:20, and 90:10), Random Forest modeling, and evaluation using accuracy, precision, recall, F1-score, confusion matrix, feature importance, ROC-AUC, and k-fold cross-validation. The 70:30 split produced 96% training accuracy and 97% testing accuracy. The testing confusion matrix correctly classified 24 of 25 GTY/PTY-appointed cases and all 5 not-appointed cases. The model also achieved an AUC of 0.992, indicating strong classification capability. These results show that the Random Forest model can provide a systematic, consistent, and objective approach to supporting teacher and staff performance evaluation and appointment decisions]]>
