mediTree

For the elective Fundamentals of Web Engineering, we were tasked with creating a visualization and editing platform for decision trees trained on medical data to extract knowledge from doctors and experiment on the viability of such models. We delivered an interactive, real-time collaborative editor that can create decision trees of variable complexity using a python microservice, supports patient simulation and in-editor experimentation and self-hosting. I was especially involved in backend development and CI/CD, which involved the creation of a sophisticated custom synchronization protocol using Y.js and Socket.io messaging.