The Informatics Engineering Undergraduate Program, Faculty of Engineering, Pelita Bangsa University, implements Outcome Based Education (OBE) by emphasizing the achievement of student Learning Outcomes (LO) as an indicator of the quality of learning in higher education. LO achievement measurement has been mostly done manually through academic assessments, so it is less than optimal in predicting student performance comprehensively. This study aims to build a prediction model for student Learning Outcomes achievement using machine learning algorithms. Research data were obtained from academic results, attendance, lecture activities, and student skill indicators. The prediction model was developed by comparing the Support Vector Machine (SVM), Random Forest, Decision Tree, and Artificial Neural Network (ANN) algorithms, with performance evaluation using accuracy, precision, recall, and F1-score metrics. The results showed that the Random Forest algorithm provided the best performance with more stable accuracy compared to other algorithms. Furthermore, the distribution of Program Learning Outcomes (PLO) in the curriculum shows: PLO 1 (57 courses), PLO 2 (10 courses), PLO 3 (3 courses), PLO 4 (27 courses), PLO 5 (8 courses), PLO 6 (20 courses), PLO 7 (33 courses), PLO 8 (10 courses), PLO 9 (54 courses), and PLO 10 (57 courses). Based on student scores in 57 courses, the distribution of assessment categories is as follows: Very Good 38.1%, Good 46.3%, Fair 8.4%, and Fail 7.2%. Thus, the PLO achievement of the Informatics Engineering Undergraduate Study Program reached 84.4% in the Good and Very Good categories. This finding provides a significant contribution to efforts to monitor and plan strategies for improving the quality of OBE-based learning adaptively and data-driven.