About Me

I'm Peerapon Wechsuwanmanee, a Machine Learning Engineer currently working at REWE Group in Cologne, Germany. I build production AI/ML systems that deliver real business value.

My career started in computational mechanics research at RWTH Aachen, where I published 10 peer-reviewed papers on damage modeling in advanced steels. The transition from simulating material failure to training ML models felt surprisingly natural - both require rigorous experimentation, statistical thinking, and the humility to let data surprise you.

Since then, I've built systems across the full ML lifecycle: from research prototypes to production deployments serving thousands of users. I've worked on computer vision for steel defect detection, Thai-language speech recognition, predictive maintenance for automotive IoT, and most recently, multi-agent GenAI systems at enterprise scale.

I care deeply about the engineering side of ML - the infrastructure, observability, and reliability work that turns a promising model into a dependable product. The best ML system is one nobody notices because it just works.

Outside of work, I teach data science and ML courses, mentor career switchers, and contribute to open-source localization efforts (including the Hugging Face NLP course Thai translation).