DoseLearn
A clinical decision-support platform for pediatric intensive care.
DoseLearn is a medical application developed in collaboration with the Pediatric Intensive Care Unit at the University of Szeged. It forms part of a broader research project focused on making long-term sedation and analgesia weaning more personalized and objective.
The application is built as a Progressive Web App with Angular and PrimeNG, providing a responsive interface for working with the project data and clinical workflows. Firebase provides the backend infrastructure, including Cloud Firestore for data storage, Cloud Functions for server-side functionality and Firebase Hosting for deployment.
The project is developed as a real-world application rather than a standalone demonstration project, which introduced requirements around reliability, data handling and maintaining a consistent development and deployment workflow.
GitLab and GitLab Runner are used to automate the development lifecycle. The CI/CD pipeline handles tasks such as building and deploying the application as well as generating changelogs, reducing the amount of repetitive manual work involved in releasing new versions.
The DoseLearn project combines frontend development, cloud services and automated delivery in a production-oriented environment while supporting an ongoing medical research initiative at the University of Szeged.
Links
$ cat challenges.md
Building for a real clinical workflow
Unlike a typical coursework application, the interface had to support an actual domain-specific workflow. Working with clinical requirements meant translating domain needs into a clear and practical user interface while keeping the implementation maintainable.
Working with a cloud-based application architecture
The application relies on several Firebase services rather than a traditional standalone backend. Coordinating Firestore data, Cloud Functions and hosting required understanding how the different services fit together and where different pieces of application logic should live.
Reliable application delivery
The project uses an automated GitLab CI/CD workflow to build and deploy new versions. This reduced manual deployment steps and made the release process more consistent and repeatable.
Automating project maintenance
The CI/CD workflow goes beyond simply building the application. It also automates tasks such as changelog generation, allowing routine release-related work to happen as part of the development workflow.
Developing a maintainable PWA
The application needed to provide a responsive experience across different devices while retaining the benefits of a Progressive Web App. This required careful component structure and consistent UI patterns across the application.
A flexible learning platform built around user-created questions.