David Salinas
Group Lead - OpenEuroLLM at Albert-Ludwigs-Universität Freiburg (Albert Ludwig University of Freiburg)
Grenoble, Auvergne-Rhône-Alpes, France
Summary
🤩
Rockstar🎓
Top SchoolDavid Salinas is an experienced machine learning researcher and leader with 11 years building production-ready forecasting and AutoML systems across academia and industry. Currently Group Lead of OpenEuroLLM at ELLIS Tübingen and a researcher at the University of Freiburg, he previously drove novel ML services and algorithms at AWS, including an open-source HPO library used by AWS services and a transfer-learning approach that improved AutoGluon Tabular performance. His work spans deep forecasting models (authoring and shipping DeepAR variants), multi-objective optimization for low-cost/low-latency ML endpoints, and probabilistic time-series dataset integration demonstrated through contributions to the widely used GluonTS project. Comfortable moving between C++ research code and Python production stacks, he combines strong theoretical grounding from a Grenoble PhD with a track record of turning research ideas into deployed systems.
11 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy (PhD) Computer Science, Doctor of Philosophy (PhD) Computer Science at Université Grenoble AlpesMaster's degree Computer Science, Master's degree Computer Science at École normale supérieure de Lyon
English, French, Spanish, German
Github Skills (12)
forecasting10
pandas10
mxnet10
machine-learning10
forecast10
time-series10
python10
data-science10
data-engineering10
numpy9
deeplearning-ai9
deep-learning9
Programming languages (4)
Github contributions (5)
awslabs/gluonts
May 2019 - Aug 2022
Probabilistic time series modeling in Python
Role in this project:
Data Scientist
Data ScientistContributions:12 commits, 17 PRs, 8 pushes in 3 years 2 months
Contributions summary:David contributed to the development of datasets for time series forecasting. They implemented access to the M4 dataset, fixed yearly data handling, and added examples to the setup.py file. Additionally, the user refactored code to correctly handle time series indices, and modified the training procedure. The commits show a focus on improving dataset integration and model evaluation for the GluonTS library.
forecastingpythontime-series-analysistimeseries-forecastingaws
geoalgo/geoalgo.github.io
Sep 2020 - Feb 2025
Build a Jekyll blog in minutes, without touching the command line.
Contributions:34 pushes in 4 years 5 months
jekyll-blogjekylljekyll-thememinutes
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.