Regardless of the spatial or temporal context, school orientation plays a pivotal role in determining the future of a high school student. Guiding the student to make the right choice in orientation requires significant effort from multiple stakeholders, including parents, friends, administration, teachers, and academic counselors. In many cases, the high school student may disregard the advice provided by these stakeholders, even though they lack sufficient knowledge to reconcile their desires with their abilities. Consequently, the final decision regarding orientation may be based on the student’s conviction, a random choice, imposed by parents, or an administrative decision. In this article, we propose a novel approach based on machine learning to predict the appropriate orientation choice for a new student based on their qualifications, results, and professional personality. In this context, we followed the standardized CRISP-DM methodology, widely used for data mining and data analysis projects. The data used were collected from students who have already undergone the orientation process. Subsequently, these data were carefully processed. During the modeling phase, we compared five algorithms, namely Decision Tree, KNN, Random Forest, SVM, and Logistic Regression. In the evaluation stage, we obtained promising results. Finally, according to our research, this approach represents an innovation in the field of secondary education, specifically in school orientation.