Summary
Eiji Nakatsugawa is a Senior Data Scientist with five years of focused experience applying machine learning to automotive ADAS/ADS perception and model-based development at Toyota and Woven by Toyota. He has taken perception models from research to mass production for global Safety Sense releases, built scalable data pipelines and evaluation systems using Kubeflow and W&B, and holds patents for data-inference search and ML-driven component modeling. His strengths span practical model compression and ECU deployment, active learning and super-resolution for object detection, and coupling ML with 1D/3D-CAE simulations to cut development cost and time. A Kaggle silver medalist in the lish-moa competition, he pairs competitive data-science practice with production-grade engineering—AWS, Git-based release workflows, and synthetic data usage. Less obvious is his cross-domain fluency: he blends physics-informed ML (PINNs), inverse RL, and statistical causal discovery to solve both perception and vehicle-systems problems. Based in Greater Tokyo, he consistently turns complex, safety-critical requirements into deployable, patent-backed solutions.
5 years of coding experience
5 years of employment as a software developer
東京工業大学
修士, 機械物理工学専攻, 修士, 機械物理工学専攻 at Tokyo institute of technology