Ahmet Erdem
Principal Data Scientist
Istanbul, Istanbul, Turkey
Summary
👤
Senior🎓
Top SchoolAhmet Erdem is a Principal Data Scientist based in Istanbul with nine years of experience building production-ready ML systems, specializing in NLP, forecasting and feature-importance methods. He has held senior roles at NVIDIA (including work on LLM agents and the logits-processor-zoo) and co-founded Epicurus Labs where he shipped consumer apps for social support and rapid opinion polling. His open-source contributions include leading development of LOFOImportance, reflecting deep expertise in model interpretability and practical evaluation. Trained at KU Leuven (MSc AI) and Boğaziçi University, he blends academic rigor with competitive machine‑learning success (KDD Cup and Kaggle golds) and a track record of optimizing models at scale using Spark, RAPIDS and C++.
8 years of coding experience
10 years of employment as a software developer
Master’s Degree Artificial Intelligence, Master’s Degree Artificial Intelligence at KU LeuvenBachelor of Science (BS) Computer Engineering, Bachelor of Science (BS) Computer Engineering at Boğaziçi University
English, Turkish
Github Skills (12)
scikit10
pandas10
machine-learning10
python10
data-science10
scikit-learn10
testing9
data-visualization8
data-visualisation8
data-visualizations8
explainable-artificial-intelligence7
github7
Programming languages (5)
Github contributions (5)
aerdem4/lofo-importance
Jan 2019 - Apr 2022
Leave One Feature Out Importance
Role in this project:
Data Scientist
Data ScientistContributions:2 releases, 13 reviews, 20 commits in 3 years 3 months
Contributions summary:Ahmet primarily contributed to the development and testing of a feature importance library, "LOFOImportance." Their initial commit set up the project, and subsequent commits added crucial features such as test cases, visualization capabilities, and support for different models, including a fast LOFO implementation. The user also worked on refactoring and improving the usability of the core library, demonstrating expertise in feature importance techniques.
feature-importanceimportancedata-sciencedeep-learningmachine-learning
aerdem4/kaggle-quora-dup
Jun 2017 - Feb 2019
Solution to Kaggle's Quora Duplicate Question Detection Competition
Contributions:11 commits, 2 PRs, 10 pushes in 1 year 8 months
kagglequestionduplicatequoramachine-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.