Pietro Berkes
VP Of Data Science
Lausanne, Vaud, Switzerland
Summary
👤
Senior🎓
Top SchoolPietro Berkes is a VP of Data Science based in Lausanne with 20 years of experience building production ML systems end-to-end, from discovery through deployment. He progressed from academic research and postdoc roles into industry leadership, founding a consultancy and later driving data science strategy and delivery at NAGRA. Pietro blends deep technical chops—contributions to flagship open-source projects like scikit-learn and joblib—with operational experience shipping robust, cache- and I/O-sensitive systems in production. Known for fixing subtle stability and cross-platform issues, he brings a pragmatic focus on reproducibility and scalable pipelines. He holds advanced degrees from ETH Zürich and Humboldt and retains an active researcher’s mindset while leading cross-functional teams.
20 years of coding experience
11 years of employment as a software developer
PhD, PhD at Humboldt-Universität zu BerlinMSc, MSc at ETH ZürichUniversity College London
English, German, French, Italian
Github Skills (28)
data-analysis10
caching10
python10
data-science10
scikit10
memory-management10
machine-learning10
parallel-computing10
scikit-learn10
pickle9
python-multiprocessing9
numpy9
multi-process9
multiprocessing9
threaded8
Programming languages (4)
Github contributions (5)
joblib/joblib
Aug 2010 - Mar 2011
Computing with Python functions.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:10 commits in 7 months
Contributions summary:Pietro primarily focused on improving the `joblib` library's internal workings and stability. They addressed pickling issues with decorated methods and Windows-specific path and file name problems. Their contributions included refactoring the `Memory` class and its interaction with disk storage and caching mechanisms, adding context managers, and fixing race conditions in test environments. The user also made several minor improvements, such as improving documentation and code style.
pythonmemoizationcachingmultiprocessingcomputing
scikit-learn/scikit-learn
Apr 2011 - Jun 2011
scikit-learn: machine learning in Python
Role in this project:Data Scientist
Data ScientistContributions:22 commits, 1 PR, 2 comments in 2 months
Contributions summary:Pietro primarily contributed to the `scikit-learn` repository by implementing and improving the `fetch_mldata` function. This function downloads and loads datasets from the mldata.org repository, integrating them into the scikit-learn ecosystem. The user focused on error checking, handling different data formats, and improving documentation, ensuring compatibility and usability for machine learning tasks.
data-analysispythonstatisticsdata-sciencelearn-machine-learning
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.