I am Matej Gazda, an AI researcher and consultant working on machine learning for medical imaging: how to build models that are accurate, calibrated, and trustworthy in clinical settings where data is scarce and the cost of being wrong is high.
Research interests
- Deep learning for medical image analysis
- Sample-efficient learning, self-supervision, and domain adaptation
- Uncertainty quantification and model calibration
- The boundary between methodological novelty and clinical usefulness
Background
I moved into AI research after several years in software engineering, where I worked on production ML systems and distributed services. That mix shows up here: most of what I write tries to be both rigorous and runnable.
Get in touch
- GitHub: BraveDistribution
- LinkedIn: matej-gazda
