Post

Master’s student of the NSU Faculty of Economics is developing an AI simulator for oral Chinese exam preparation

Master’s student of the NSU Faculty of Economics is developing an AI simulator for oral Chinese exam preparation

Published on: 2026-08-24

Source: Novosibirsk State University –

An important disclaimer is at the bottom of this article.

Master’s student of the Faculty of Economics (FE) at Novosibirsk State University Mikhail Savin with the support of the NSU Startup Studio, became the winner of the “Student Startup” competition with the project HSK Chat — a web application for preparing for the oral part of the HSKK Chinese language exam. The development received a grant from the Foundation for Assistance to Innovations and is aimed at creating an affordable tool for regular language practice.

In 2026, NSU set a record for the number of winners in the “Student Startup” competition: 26 student projects from the university received grant support, which is twice the number from the previous year. Among the winners, the largest number of projects — six — were presented by students from the Faculty of Economics.

There are applications on the market for studying words and characters, platforms with grammar and tests, HSK preparation courses, as well as separate tools for practicing spoken Chinese. However, according to the developers, their solution differs by focusing specifically on the oral part of the exam.

We combine several elements in one service: exam practice tasks, oral response recording, automated result processing, recommendations, and preparation history preservation. Our mission — to make preparation for HSKK more accessible, regular, and convenient, without the need to constantly work with a teacher, — says Mikhail Savin.

The key technical element of the application is working with the user’s spoken speech. The user opens a training task, as close as possible to the exam format, and records their response through a microphone. The audio recording of the answer is processed using artificial intelligence technologies and automated speech recognition, after which the system generates feedback. At the same time, the developers do not aim to completely replace the teacher with artificial intelligence.

Our task is to provide the user with quick automated feedback for self-training, — emphasizes Mikhail.

The idea for the project was born in a student course at the SMBA Faculty of Economics, where the future creators of HSK Chat met and decided to create a project together that could change the approach to learning the Chinese language.

We know how difficult it is to learn Chinese, especially to speak it fluently. And we thought: if technology can already recognize speech, analyze data, and provide feedback to a person — why not use this to help them start speaking Chinese? We are starting with HSKK, but we are looking much further — we want to create a new technological foundation for language exam preparation, where artificial intelligence becomes a personal coach for each individual. Our goal is not just to help pass the exam. Our goal is to change the very approach to language learning., shares Mikhail Savin.

A team of four people is working on the project, each responsible for their own area:

  • Mikhail Savin, first-year master’s student at the Faculty of Economics, NSU, CEO of the project. Responsible for overall management, strategy, project development, and key decisions. Has over five years of experience in launching tech platforms and startups, oversees the technical part and product architecture.
  • Elizaveta Kuzmicheva, a 4th-year undergraduate student at the Economics Faculty of NSU, product manager and lawyer of the project. She is responsible for product strategy, user scenarios, team coordination, and legal support.
  • Jovan Vid Minovich, a first-year master’s student at the NSU Faculty of Mechanics and Mathematics, frontend/fullstack developer. Responsible for platform development, interface, integration with the server part and API, as well as system performance.
  • Vitaliy Lukash, a second-year undergraduate student in the “Regional Studies” program at NSTU, a methodologist and task specialist. He is responsible for the content of the product: tasks, methodological materials, and their compliance with the HSKK format.

The development has been underway for about six months. To date, a prototype of the web application has been created — this is no longer just an idea or concept, but a product ready for demonstration. The application is being developed as a web service accessible through a browser both from a computer and a mobile device. The architecture includes a user interface, a server part, a database, an audio recording module, and a results processing module.

The grant funds are planned to be directed towards the development of the software component, creation and filling of educational content, methodological support, project management, and legal support. The team intends to implement basic functionality: registration and personal account, a task database, recording of oral responses, automated processing of results, recommendations, and preparation history. After creating the first working version, user testing, feedback collection, and product refinement are planned.

Primarily, the product is intended for the field of digital language education. The main users are people learning Chinese and preparing for the HSKK. The service can also be used by language schools, teachers, educational organizations, and learning centers as an additional tool for home practice.

— Our service allows part of such practice to be transferred to a digital format: a person can independently open an assignment, complete it, record the answer, and receive feedback, — concludes Mikhail Savin.

The project is supported by the Foundation for Assistance to Small Innovative Enterprises under the program “Student Startup” of the event “University Technology Entrepreneurship Platform” of the federal project “Technologies.”

Please be informed; This information is raw content obtained directly from the information source. It represents an accurate report of what the source claims and does not necessarily reflect the position of MIL-OSI or its clients.