The management of the KursAnt group of companies came to us with a request to develop an IT platform for testing school students according to UNT standards. At that time, 2 private schools and an educational UNT preparation center were already operating under the KursAnt brand. Offline, they taught more than 3500 students and held testing every week, mostly on paper.

As the scale grew, this became a serious problem. The process took a lot of time, required many resources and was inconvenient for both teachers and students.

They tried different solutions on the market, but none fully covered the tasks: somewhere the question base was outdated, somewhere the system was unstable, and somewhere the product simply did not match the latest UNT requirements.

At some point, it became clear that creating their own system would be more effective, first for internal use and later as a separate product for the B2B market.

After fixing the client’s vision and tasks, we deeply immersed ourselves in the exam logic, studied standards, question types, test structure and all nuances that affect the final result.

The key task was to make a flexible tool for methodologists. They needed not just a “set of tests,” but a full constructor through which exams could be configured for different scenarios.

We implemented a panel where a methodologist can assemble an exam like a constructor: define required and optional subjects, set the number of questions, distribute them by difficulty levels, topics and formats. For example, specify what percentage of tasks should be simple and what percentage should have multiple answer options.

As a result, the platform can flexibly configure different exam formats. In total, we implemented about 15 types of tests, question types and passing states.

We separately strengthened the system with analytics. We already had a base of more than 20,000 questions across all subjects and topics from grades 5 to 11, and we proposed using this data volume for deeper analysis.

The platform gained functionality that allows not just seeing a student’s result, but analyzing their knowledge more deeply: where they make mistakes, which topics are weak and how dynamics change.

Because of this, the system became not only a testing tool, but also a full base for analyzing and improving student results.

After launch, we ran several waves of testing with real school students to collect feedback and understand how the system works in a live environment. It was important for us not just to “launch,” but to make sure the product really solves the problem.

As it turned out, for students this became more than just a testing system. The platform began acting as a personal analyst: showing weak spots, helping track progress and giving an understanding of what needs work.

An engagement element was also added: ranking, competition with friends and result comparison. This made preparation not only useful, but also more interesting.

For teachers, the system simplified control and analysis. Tests no longer need to be checked manually and data no longer has to be collected by hand; everything is visible in the system with clear analytics for each student and group.

It is especially nice that the product started moving beyond internal use. Dozens of private schools connected to the system in B2B format, confirming that the solution is demanded by the market and solves a real pain.

In the end, this is not just a testing tool, but a full educational platform that can be scaled and developed further.