Translating ML models from research into clinical practice — from code to clinic — requires infrastructure that manages data governance, reproducible training and reliable deployment under strict regulatory requirements.
The Institute for AI in Medicine (IKIM) at University Hospital Essen is mandated to develop, evaluate, and deploy algorithms within the clinical setup. To address the challenges of this highly regulated environment, we have built an open-source Internal Developer Platform based on the idea of the BACK stack—combining GitOps, declarative infrastructure, and policy-driven governance.
The platform enforces compliance through Golden Paths—predefined workflows that standardize the development lifecycle from data collection and curation to training and validation, while giving data scientists self-service access to infrastructure.
This talk covers practical patterns for embedding compliance into the developer experience — in a domain where every shortcut is a systemic risk.