Interview with George Kaissis

Github: @gkaissis  |  Twitter: @gkaissis

Where are you based?

"Munich, Germany"

What do you do?

"I’m an assistant professor at the Technical University of Munich where I work at the Institute for Artificial Intelligence in Healthcare and the Institute of Radiology. I’m also a Research Associate at the Department of Computing at Imperial College London and I lead the OpenMined Healthcare Unit."

What are your specialties?

"I’m a specialist diagnostic radiologist and computer scientist. My research focuses on privacy-preserving medical imaging analysis, specifically on differentially private deep learning and decentralized collaborative learning workflows as well as Bayesian and probabilistic machine learning methods."

How and when did you originally come across OpenMined?

"I joined OpenMined as a research scientist in 2020 and was welcomed very warmly by many great researchers and friends."

What was the first thing you started working on within OpenMined?

"I started a project on secure and private deep learning in medical imaging which eventually led to the creation of our library called PriMIA ( PriMIA was developed by Alex Ziller, who’s now also an OpenMined research scientist with help from many OpenMined project members and is a federated learning library using PySyft and PyGrid. It allows users to leverage differentially private federated learning with secure multi-party computation for end-to-end privacy preserving deep learning on medical images."

And what are you working on now?

"I now lead the OpenMined Healthcare Unit which is the central hub for healthcare-related research, engineering, project management and partner outreach at OpenMined. Alongside this, I’m also a member of the OpenMined steering committee and continue to pursue academic research in the field of privacy-preserving machine learning."

What would you say to someone who wants to start contributing?

"OpenMined is one of the most diverse, welcoming and inclusive communities I have ever been part of. There is always a helping hand and we have many great resources to get you started! Also if you are interested in privacy-enhancing technologies but feel your research or work is not directly related, come hang out anyway! You will likely be surprised how applicable questions of information flow, transparency and privacy are in all areas of science and society!"

Please recommend one interesting book, podcast or resource to the OpenMined community.

"If you like reading books, you can’t go wrong with The Ethical Algorithm by Michael Kearns and Aaron Roth. If watching videos is your thing, I’d definitely recommend our OpenMined course!"