A student at Novosibirsk State Technical University, (NETI), has developed the "OPD Assistant", an intelligent chatbot that helps students master the disciplines "Fundamentals of Project Activity" and "Project Activity". It answers questions about the course, helps to generate ideas for projects, explains difficult points and can be adapted to work in other universities.
The development has already been highly appreciated by experts, becoming the winner in the "Best Student Project" nomination at the NSTU Educational Session in 2025 and taking first place in the "IT Project" nomination at the Student Project Competition in December 2025.
The bot is based on a modern large language model, which has been specially adapted to effectively answer questions in disciplines based on an accumulated array of questions and answers. "This process is called RAG (Retrieval-Augmented Generation — extraction with extended generation), when the system does not just communicate based on the "shared memory", but precisely searches for answers in the knowledge base provided to it — answers from teachers," said Egor Antonyants, project manager, assistant professor at the Department of Automated Control Systems at NSTU-NETI.
The result of the development is a digital assistant that can provide students with a clear and structured course response at any time. "We analyzed hundreds of sample questions from students from previous years and configured the bot so that it understands the essence of the request, even if it is not formulated academically, but in simple language," said Zakhar Silantyev, one of the project developers, a fourth—year student at the Faculty of Automation and Computer Engineering.
According to Egor Antonyanets, educational chatbots are becoming a trend in universities around the world. However, many of them are either simple FAQ databases or use "raw" language models without deep adaptation to a particular discipline. The "OPD Assistant" of NSTU-NETI is distinguished by its deep specialization for a unique learning process, which makes its answers more relevant and useful. In addition, the development solves the acute problem of mass education — the lack of individual attention in large streams. The prototype is already capable of processing up to 500 requests per day with an average response time of less than 5 seconds. The introduction of such an assistant makes it possible to reduce the flow of standard questions to teachers by up to 75%, which frees up time for deeper work with students.
The project was created specifically by order of Natalia Plotnikova, Head of the Department of Education Strategy at NSTU-NETI, to support the educational process in disciplines related to project activities. The development has a huge potential for scaling: the technology can be adapted for any discipline or university, acting as a digital tutor for massive online courses, an internal help system for adapting first-year students, and a support tool for students working on course and graduation projects.