KrokMed
A platform where Ukrainian medical students could complete mock tests for their national KROK exams. Co-founded this platform, built every technical part of it, and supervised it while active, managing revenue and costs.
The problem
The main government website for exam preparation could not handle an extensive traffic increase, especially weeks before exams, and overall it struggled for quite some time, at the same time providing an outdated and laggy interface. This meant students were left without any means to prepare for exams, which was crucial, since a handful of similar or identical questions appear on actual exams and it is worthwhile completing mock ones.
The product
The main goal was not an innovative design, but rather the clear functionality users required. We settled on a clean background and a minimalist design. Straightforward filtering and a test library were built. What was innovative was the ability to skim through the whole question bank in a row, letting students get familiar faster with each question that might repeat on an actual exam.
The test catalogue.
Under the hood
As for the backend, FastAPI was used, with plain HTML, CSS and JavaScript on the frontend. The database was built on PostgreSQL, and the website was hosted in the cloud on its own domain. Email confirmation services — such as the message sent to notify a user of a successful payment or registration — were managed separately. The database tracked users, active sessions per device and which user bought access to which test. All development, plus the management of revenue and costs, was my side of the work, while my partner was in charge of branding, marketing, and providing the website with mock exam questions.
An example of an exam question.
The numbers
Hundreds of website visits a day at peak, and 67 successful sales in total — two of which were development test payments and were refunded, leaving 65 real customers. That comes to roughly 15,000 UAH, about 300 EUR, in revenue.
The payment dashboard.
Why it closed and what it taught me
The main problem with this project was the faded demand outside the exam period. Not only were revenues too small during the quiet periods to cover the costs, but by the time the next season came around, more affordable and even free solutions had started to arrive.
The deeper problem was a low customer conversion rate, together with a sudden halt in developing new features that would keep people choosing our paid option over free ones with less functionality. A sudden dopamine surge from small success, and the absence of a clear growth path, stopped any new ideas from being implemented. In conclusion, this project needed more discipline, planning and marketing effort to retain users and grow. That is the lesson I will carry into all my future projects.