josefondrej / maml

Code for "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"

Hello! 👋

My name is Josef Ondrej and I enjoy building useful mathematical and machine learning models.

I have a solid background in mathematics ➗, particularly in statistics and probability theory, from my studies at the Charles University in Prague. But besides solving math problems I also certainly do not shy away from scraping data, digging into it and writing well tested ✅ production code with CI/CD pipelines and monitoring 🖥 -- skills that I gained while developing and operating the Watson Assistant Autolearn feature at IBM.

I also spent a couple of years promoting the use of MLOps at sports betting company ⚾ Tipsport mainly using Jenkins, Docker, OpenShift, ArgoCD, MLFlow and Helm Charts while also developing models for predicting sports matches outcomes in Tensorflow and PyTorch.

My language of choice is Python 🐍 (in which I developed a few small libraries and projects like Snax, Transplants, SVGecko, API-Watchdog and GlucoScan) but I have also worked a lot with Java and have some experience also with R, JavaScript, Groovy (for Jenkins plugins), C# and a few others. I can also manage some basic web development (React, HTML, CSS) if needed.

In my free time I enjoy climbing and hiking in the mountains ⛰️.

For more information you can check out my CV or website or feel free to contact me.