This research proposed and developed a computational framework for such scenarios using Machine Learning (ML) techniques to help predict the most suitable levels for students when transferring between curricula, assigning these levels automatically, and holding students' data throughout their academic journey. Students' datasets were collected from their educational records for two consecutive academic years to fulfil this goal, and then pre-processing techniques were applied to the raw dataset. The research focused on how machine learning can predict students' levels using several models including Artificial Neural Network and Random Forest, alongside assembled classifiers. Extensive simulation results indicated that the Levenberg-Marquardt Neural Network method (LEVNN) has the best average results among the other applied methods. A user-friendly platform has been designed based on a web-based student management system to bring both perspectives together in one platform for schools and parents. The research would help education providers predict students' correct levels more efficiently without regular examinations, saving time and cost for schools, students, and parents.