On-time graduation is one of the key indicators of academic quality that affects study program accreditation and institutional reputation. Many Informatics Engineering students, however, still complete their studies later than scheduled due to various academic and non-academic factors. This study aims to apply the Random Forest algorithm to build a predictive system that classifies students into “on-time” and “delayed” graduation categories based on academic data such as cumulative grade point average (GPA), number of credits completed per semester, attendance rate, and final project completion status. The research method follows the CRISP-DM framework, covering business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The model was trained using an 80:20 train-test split and evaluated using a confusion matrix along with accuracy, precision, recall, and F1-score metrics. The results show that the Random Forest algorithm achieves a good level of accuracy in classifying student graduation status, with GPA and attendance rate identified as the most influential variables. This model has the potential to be used by study programs as an early-warning system to identify students at risk of delayed graduation so that academic intervention can be carried out sooner.