The main web application of the medical center smt-clinic.ru, built with the Next.js and Django REST Framework stack. It features a functional admin panel, SEO optimization, direct integration of doctors' schedules, and a user-friendly appointment booking system.
The application’s front end is built with Next.js, React 19, and Zustand. It leverages server components and an FSD architecture integrated into Next.js’s route-based framework. Rather than using heavy UX libraries, a custom lightweight UX was developed, resulting in minimal bundle sizes, exceptional performance, and strong SEO optimization — pages load almost instantly.
Dark and light themes are available, along with adjustable font sizes for visually impaired users. The app is fully responsive and optimized for seamless display on any screen size. Given the medical center’s prominence and reputation in the region, service quality and universal accessibility—including for users with disabilities—are top priorities.
The application is integrated with ChatGPT via my custom proxy service IMVO.SITE. Thanks to this setup, a “patient assistant” has been implemented—an intelligent site-wide search that responds to user queries with contextual commentary and recommendations.
In addition to the features previously described, other AI-based solutions have been implemented in the service. The administrative panel now includes tools for automatically generating initial content based on titles. This significantly speeds up the work of the clinic's content managers. Furthermore, some sections, such as price list items, are automatically supplemented with auxiliary descriptions. This approach enhances the user experience, helping patients better understand complex medical services. It is important to note that without automation, the process of populating large volumes of content, especially in the price list, would be very costly for the clinic.
The application features direct patient appointment booking via the website. An independent FastAPI microservice handles available slot parsing and the booking process, interfacing with the SMT Clinic HIS and a message broker to synchronize with the CRM. Both components are documented as separate case studies, highlighting their autonomy within the overall system.
Our project's backend is built on a time-tested and reliable synchronous stack, comprising Django, Redis, Celery, and PostgreSQL.
The choice of Django was driven by several key factors:
select_related and prefetch_related).It's important to emphasize that the use of a synchronous stack in this context is a deliberate and justified decision. The main volume of requests from the client-side is handled by a dedicated Node.js backend (BFF) for the Next.js frontend. Meanwhile, the most resource-intensive real-time operations are delegated to specialized asynchronous services, ensuring the responsiveness and performance of the main system.
If necessary, I can involve designers, data analysts, and senior-level developers in the project