A Russian Researcher Built an AI That Can Predict Student Performance — and Teach Them Too
In southern Russia, a computer scientist is developing a neural network that can forecast grades, identify knowledge gaps, and even act as a virtual teacher.
In the republic of Dagestan, Tagirbek Aslanov, an associate professor of information security, has created an artificial intelligence system designed to transform how students learn. His neural network can predict a student’s academic performance for the following school year with just a 4% margin of error.
The algorithm analyzes each child’s learning patterns, tracking how different stages of education affect results. “It’s not just about grading,” Aslanov explains in local media. “It’s about understanding how every student learns — and adapting to that process.”
A Virtual Teacher in the Making
Aslanov’s system is part of a broader initiative called “Virtual Teacher,” a digital learning platform designed to replicate the functions of a real educator. The AI can answer students’ questions, conduct oral quizzes, grade written assignments, and monitor engagement and emotional state during class.
Another tool, “Knowledge Graph,” maps a student’s understanding of different subjects to pinpoint missing concepts — data that helps build a personalized learning path for each child.
Building a Smarter Education System
The project reflects a larger national push to modernize education through AI and digital services. Russia’s “My School” online platform already connects millions of teachers, students, and parents, combining an electronic gradebook, a vast digital library, and teaching materials in one system.
Aslanov’s work takes that vision a step further — toward classrooms where every child learns at their own pace, guided by algorithms that know their strengths and weaknesses better than any report card.
