# ./moritz.md

Moritz Stammel

A short README. Last edited: May 2026.

I work somewhere in the triangle of research, engineering, and policy, mostly around machine learning for medical imaging and its use in practice. I prefer the parts of the problem where these three meet, even when they disagree.

// the triangle, roughly to scale

  • research asks the right questions
  • engineering makes the answers work
  • policy keeps the consequences in view

I am particularly interested in how machine learning systems behave when they leave controlled settings and have to work under real-world constraints.

research

At Imperial College London and previously TU Dresden, I have worked on problems in medical imaging, including representation learning, model robustness, and counterfactual approaches for stress-testing models.

My recent work focuses on test-time counterfactuals for evaluating model behavior under distribution shift.

engineering

In my work with QuantCo and Virdx, I build and evaluate machine learning systems in clinical settings, where model behavior has to remain reliable under real-world constraints, the parts that have to keep working on a Tuesday morning.

At Virdx, I work on machine learning for medical imaging, with a focus on making non-invasive imaging a more reliable basis for diagnosis.

policy

Through my work on the executive board of the Studierendenforum im Tönissteiner Kreis and the Rethinking Healthcare initiative, I work on questions around the governance and deployment of AI in healthcare.

elsewhere

<!-- thanks for reading. -->