Patrick Rockenschaub
Group Leader at Medizinische Universität Innsbruck
Innsbruck, Tyrol, Austria
Summary
🤩
Rockstar🎓
Top SchoolPatrick Rockenschaub is a research group leader and machine learning scientist with 10 years’ experience developing clinical decision-support models for intensive care using heterogeneous EHR time-series. He builds foundation models that generalise across hospitals and populations to enable real-time ICU decisions, and is PI of the €1.5M ERC Starting Grant project GPT-MEDIC beginning in 2026. His background spans academic and industry roles—Humboldt Fellow at Charité, Senior Researcher at Fraunhofer IKS, and senior statistician roles—combining rigorous health-data science (PhD UCL) with applied trustworthy-AI research. Patrick has hands-on experience contributing to open-source time-series tooling (sktime) by implementing PCA transformers for univariate series, reflecting a practical bent for robust, reusable methodology. He leads federated ICU data initiatives (INDICATE) and hires PhD students interested in clinical ML, foundation models, and ICU analytics. Based in Innsbruck, he balances statistical rigor with systems-minded engineering to translate complex clinical data into deployable decision support.
10 years of coding experience
2 years of employment as a software developer
Exchange Semester at the College of Business Administration Economics, Exchange Semester at the College of Business Administration Economics at Seoul National UniversityBachelor’s Degree Economics, Bachelor’s Degree Economics at WU (Vienna University of Economics and Business)University College London
English, German
Github Skills (14)
scikit10
scikit-learn10
machine-learning10
pca10
time-series10
python10
data-science10
data-analysis10
numpy9
pandas9
unit-testing8
ai7
sktime7
data-mining7
Programming languages (4)
Github contributions (5)
sktime/sktime
Oct 2019 - May 2020
A unified framework for machine learning with time series
Role in this project:
Data Scientist
Data ScientistContributions:14 commits, 9 PRs, 1 push in 7 months
Contributions summary:Patrick contributed to the sktime library by implementing and testing a Principal Component Analysis (PCA) transformer for univariate time series data. Their work involved adding a wrapper around the `sklearn.decomposition.PCA` class, including associated tests to ensure correct functionality and output format. The user also addressed code issues and merged changes from a related branch.
forecastingtime-series-analysistime-series-regressiondata-sciencedeep-learning
prockenschaub/icuDG
Jul 2022 - Mar 2023
Domain Generalisation for ICU risk prediction models
Contributions:190 commits, 2 PRs, 54 pushes in 7 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.