The article describes the issue of modeling the digital twin of a teacher in the context of the task of forming individual educational tasks for students. An approach is proposed that analyzes the patterns in the teacher's choice of the most priority topics and areas for each student based on the available data on his previous experience. The article deals with the search for patterns between the digital twins of teachers for the subsequent prediction of their choice of topics and sections of training materials for the individual work of students. As the basis for creating a dataset for the formation of digital twins of teachers, the methodology of decomposition of tasks by tags is used (tags are designation of typical tasks within the framework of tasks by students). The paper describes the relationship between the student's digital trace, the digital trace and the teacher's digital twin as a part of solving the problems of managing the process of forming individual tasks.