Kohei Ozaki
Fellow at 株式会社Rist
Japan
Summary
🤩
Rockstar🎓
Top SchoolKohei Ozaki is a Fellow and seasoned software engineer with 17 years of experience specializing in machine learning, data science, and scalable production systems. He blends deep algorithmic instincts—honed through ACM/ICPC, ICFP contests and a KDD Cup 2015 first-place win—with practical engineering across startups and research-driven companies like Preferred Networks and Recruit Technologies. His hands-on work spans fraud detection, recommender systems, advertising ML, and algorithmic trading, and he has contributed to notable open-source ML tooling such as Optuna (improving LightGBM tuning) and quality-assured Hivemall tests. Comfortable moving between R&D and product delivery, Kohei has also advised initiatives for industry and government AI contests and led technical efforts as a founder and fellow. Based in Japan, he holds graduate degrees from Nara Institute of Science and Technology and pairs academic rigor with a track record of top-tier Kaggle performance and system-level impact.
17 years of coding experience
8 years of employment as a software developer
D Eng, Information Science, D Eng, Information Science at Nara Institute of Science and Technology
Github Skills (10)
hyperparameter-optimization10
javas10
machine-learning10
junit10
python10
java10
lightgbm10
optuna10
testing10
apache-hive8
Programming languages (9)
Github contributions (5)
optuna/optuna
Aug 2019 - May 2020
A hyperparameter optimization framework
Role in this project:
ML Engineer
ML EngineerContributions:2 reviews, 84 commits, 34 PRs in 8 months
Contributions summary:Kohei contributed to the hyperparameter optimization framework, Optuna, by focusing on integrating and improving the LightGBM tuner. Their work involved removing an older sklearn API, fixing a bug related to tuning the `bagging_fraction` parameter, and minor updates to the codebase, including adding type hints. The user's primary contributions involved improving the functionality and integration of LightGBM within the Optuna framework.
pythonoptimization-frameworkparallelhyperparameteroptimization
myui/hivemall
Oct 2013 - Oct 2013
Scalable machine learning library for Apache Hive/Spark/Pig
Role in this project:QA Engineer / Test Automation Engineer
QA Engineer / Test Automation EngineerContributions:9 commits in 2 days
Contributions summary:Kohei primarily focused on testing and validating the `hivemall` library. Their commits added and modified JUnit tests for various machine learning UDFs, including those related to hashing and the Passive Aggressive algorithm. They ensured the functionality of these UDFs by creating test cases that covered different input scenarios and parameters. The user's work is critical for maintaining the quality and reliability of the machine learning functionalities within the library.
pigscalableapachemachine-learningbig-data
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.