currently pursuing information systems @ national university of singapore. particularly interested in backend development and making difficult things work! in my free time, i also enjoy playing football, gaming and trying different types of coffee :)
- Built an end-to-end ML pipeline using XGBoost trained on 25 seasons of match data
- Achieved ~65% outcome accuracy, RMSE ~1.0 for goal predictions, and Brier Score of ~0.40
- Engineered 30+ features including historical performance, team strength, and fixture dynamics
- React + Next.js frontend, RESTful backend using OpenAI Assistant API
- Implemented Retrieval-Augmented Generation with vector embeddings for smarter responses
- Used streaming to enhance interaction speed and responsiveness
📫 let's connect: LinkedIn | 📩 javieryeow@gmail.com