Franz Király
Director at sktime
Germany
Summary
🤩
Rockstar🎓
Top SchoolFranz Király is a Director and open-source AI leader based in Germany, focused on open research, technology sovereignty, and democratic technology governance. He leads the German Center for Open Source AI and is a core developer and council member of sktime, contributing to time-series forecasting tooling and test infrastructure, including work that touches PyTorch forecasting components. His career bridges academia and industry—former lecturer at UCL, Turing and Oberwolfach fellow, and principal data scientist roles at Shell and GfK—bringing research rigor to production challenges. Unusually, he holds advanced degrees spanning physics, mathematics, medicine and computer science, which informs an interdisciplinary approach to AI systems. Known for pragmatic engineering craftsmanship, Franz emphasizes maintainability and reproducibility in open-source software while helping public and private stakeholders integrate and govern AI responsibly.
8 years of coding experience
10 years of employment as a software developer
Dipl. phys. (BSc & MSc equivalent German degree), Physics (Physik Diplom), Dipl. phys. (BSc & MSc equivalent German degree), Physics (Physik Diplom) at Ulm University
Github Skills (18)
dataanalysis10
timeseries10
pytorch10
pytest10
python10
machine-learning10
time-series10
blackboard10
time-series-forecasting10
sktime10
test-automation10
data-analysis10
statistical-modeling9
forecasting9
statistical-modelling9
Programming languages (12)
Github contributions (5)
sktime/sktime
Mar 2019 - Jan 2023
A unified framework for machine learning with time series
Role in this project:
Backend Developer & Test Automation Engineer
Backend Developer & Test Automation EngineerContributions:86 releases, 6151 reviews, 2239 commits in 3 years 10 months
Contributions summary:Franz's commits primarily focused on enhancing the test suite, particularly by expanding the scope of test execution to cover more components and configurations. They made code changes across various test files, including those for model evaluation, transformations, and annotation. In addition, the user's work involved addressing inconsistencies in test practices, and refactoring the test infrastructure by introducing more robust or convenient patterns for testing and validation.
forecastingtime-series-analysistime-series-regressiondata-sciencedeep-learning
sktime/pytorch-forecasting
Aug 2024 - Mar 2025
Time series forecasting with PyTorch
Role in this project:Back-end Developer
Back-end DeveloperContributions:5 releases, 63 reviews, 138 PRs in 6 months
Contributions summary:Franz's commits primarily involve linting and code formatting changes using `black`. They've modified several Python files, including those related to data encoding, metrics, model implementations (Temporal Fusion Transformer, DeepAR, NHiTS), and testing. These changes suggest a focus on code style consistency and maintainability within the project. The edits touch different areas of the project suggesting the user is working on core functions.
forecastingpythonartifical-intelligensetimeseries-forecastingpytorch-lightning
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.