The automation of the educational process and the introduction of adaptive learning are becoming the most effective ways to solve the problem of the shortage of highly qualified personnel under digitalization of education.The article proposes an approach to the formation of individual educational trajectories as a composition of pedagogical experiments, which is characterized by good interpretability in comparison with methods based on machine learning. This approach, unlike most modern parametric solutions, involves an experimental search for an individual educational trajectory for each student of the group, and not just a part of the students of the group in which the application of a specific parametric solution is effective. It is proposed that the procedure of selecting educational technologies be automated similarly to how it is done by teachers in small groups, for example, by applying the automated trial and error method and its optimized versions. The possibility of transferring educational technologies to other courses is considered, without which technology reuse and scaling are impossible. A new function of the learning system is described, i. e. the formation of an individual educational trajectory, which is used throughout the study in higher education.The obtained results can be used in the construction of modern systems of individualization of learning and conducting distributed pedagogical experiments, which, in turn, reduces the threshold of teachers’ entry into the study of new learning technologies and accelerates the widespread dissemination of promising educational technologies.