This paper is focused on monitoring the current academic performance of students using artificial intelligence. Artificial intelligence technologies and, in particular, neural networks have gained wide popularity in the field of education, as they make the learning process the most effective. The use of artificial intelligence allows to automate and optimize educational processes, mainly in the issue of monitoring the academic progress of students. Monitoring the current academic progress of students is necessary, first of all, in order to timely track failing students, identify their problem areas and provide timely help and support. In order to monitor students’ academic performance and diagnose their learning problems in a timely manner, a neural network was used. The neural network dealt with the digital twins of the students. Digital twin refers to a set of various parameters of students, including grades, activity in classes, timeliness of handing in practical assignments and tests, and absences. After training, the neural network is able to identify students who are at risk, that is, the digital twin of the considered student matches one or more digital twins of students from previous years who were expelled. The performed research is of great value for all participants of the educational process, as it facilitates and automates the work of teachers, and students receive timely help, support and motivation to continue their studies in university. #COMESYSO1120.