In the Dr.Max project, we took part in an AI experiment designed to test how far the development of a real application could be pushed with the help of artificial intelligence. The brief was clear: create a functional reservation system for pharmacy-based health screenings that could serve as a shared internal solution for multiple countries.
Dr.Max already allows customers in several markets to book selected pharmacy services and health screenings, such as CRP tests, glycated hemoglobin tests or cholesterol level measurements. Each country, however, has been working with its own solution. The experiment therefore opened up an opportunity to prepare an application that could be developed centrally and adapted to local needs.
Development with Claude Code
The main idea behind the experiment was simple: build the entire application through prompting in Claude Code. All changes were created with AI assistance, while the developers’ role was to define requirements, review outputs, guide the solution and keep the project moving in the right technical direction.
The experiment was handled by a three-member supplier team. Rostislav Štěpina from MageXo used AI to prepare the business analysis, functional specification and user stories. My colleague Viktor Novák focused on the backend and application logic, while I prepared the frontend.
Go backend as the core of the system
The backend service was built in Go and formed the core of the entire reservation system. It handled the generation of time slots for screenings and pharmacies, reservation creation, the reservation flow, as well as email and SMS notifications sent to customers.
Working with AI quickly showed us that the quality of the output depends mainly on the quality of the input. It was important to define the project structure, file-handling rules, expected application behavior and the boundaries within which the AI should operate. With this initial setup in place, Claude Code was able to generate solid code that could be further developed.
A customer-facing web app and administration for pharmacists
The frontend part consisted of a web application where customers can create a reservation for a health screening. The solution also included an administration interface for pharmacists, where they can see an overview of reservations, edit reservation details and manage selected information about screenings.
The goal was to prepare a functional flow from selecting a screening to managing the reservation in the pharmacy. This allowed the prototype to be tested very quickly on specific scenarios and presented as a complete solution
Iterations, testing and demo
We moved from the first concept to a functional application in approximately 3 weeks. Development was iterative: AI prepared solution proposals, while the team continuously reviewed, adjusted and expanded them according to the needs of the application.
We then went through the entire reservation flow, from creating time slots and saving reservations to displaying data in the administration interface and sending customer notifications. The result was presented to the Dr.Max management team as a functional prototype that demonstrated a usable foundation for a real system built with an AI-first approach.
What we validated
The result of the experiment was a functional reservation system with a customer-facing web app, administration for pharmacists, notifications and integration with the existing Dr.Max environment. In other words, we created an application that came close to real deployment and showed us what AI-first development can look like in practice.
The initial setup proved to be especially important. When AI has strong context, clear project instructions and a team that can continuously evaluate its proposals, it can significantly accelerate development.
At the same time, the experiment showed that in an AI-first approach, the role of the developer shifts more toward accurate prompting, quality control, architectural decision-making and continuous guidance of the next steps.
First deployment in Slovakia
Slovakia became the first country to start using the new system. The reservation application is now live, allowing customers to book selected health screenings in Dr.Max pharmacies. You can access it at https://rezervacie.drmax.sk/.
For us, this is the most important part of the entire experiment. We did not validate the AI-first approach only on a prototype, but on a solution that moved into real use.
Frequently Asked Questions
What does AI-first application development mean?
AI-first development means that artificial intelligence is actively used during the design,
analysis and programming of an application. Developers continuously review outputs,
set the technical direction and make decisions about the quality of the solution.
What role did Claude Code play in the development of the reservation system?
Claude Code helped generate parts of the application based on prompts and project
context. The team used it when building the backend, frontend and application logic,
while the outputs were reviewed and further refined by the developers.
Can AI be used to create a real application ready for use?
Yes. With the right brief, strong technical guidance and continuous quality control,
AI can significantly speed up the development of a real application. In this case,
the reservation system moved from prototype to real use in Slovakia.
How long did it take to develop the Dr.Max reservation system?
The team moved from the first concept to a functional application in approximately 3 weeks.
During this time, the reservation flow, administration and notifications were prepared,
tested and refined iteratively.
What features does the Dr.Max reservation system include?
The system allows customers to book selected health screenings in Dr.Max pharmacies.
The solution also includes administration for pharmacists, reservation management,
time slot generation and email and SMS notifications.
What backend was used to develop the system?
The backend service was built in Go. It mainly handles time slot generation,
reservation creation, the reservation flow and customer notifications.
Does AI replace the developer in this type of development?
AI does not replace the developer. It changes the developer’s role. The developer focuses more on the brief,
architecture, quality control, decision-making and guiding the next steps.
Technical responsibility remains with the team.
Why is a good initial setup important in AI development?
AI needs clear context, project rules, technical constraints and precisely defined
expectations. The better these inputs are prepared, the stronger the outputs it can generate.
Where is the Dr.Max reservation application available?
Slovakia was the first country to start using the new reservation system.
Customers can book selected health screenings through
rezervacie.drmax.sk.
What did this AI experiment show in practice?
The experiment showed that AI can significantly accelerate the development of digital products
when the team has a strong brief, technical oversight and a clearly defined process.
It also confirmed that the AI-first approach makes sense for solutions
that are intended for real-world use.
