Brandon Ma

I’m a Data Scientist who turns complex operational problems into practical, production-ready systems. At Gildan Activewear, I supported a smart application using MIP optimization to create optimal loads based on customer needs.

I built an end-to-end application to optimize scheduling for dyeing and cutting by modeling it as a multi-vehicle routing problem. I also refactored production Jupyter Notebook code into maintainable, production-ready scripts using software engineering best practices.

I develop models that improve quality and reliability, including an anomaly detection model to pinpoint dyeing machines using abnormal water usage. I also implemented neural networks to detect when fabric would get tangled during the dyeing process.

Previously, as a Software Developer Intern at Genetec, I contributed to internal tools and improved stability and user experience for the AutoVu platform by resolving UI bugs and building frontend enhancements with Angular. I’m also proud of my projects like a Stock Mention Detection System using Whisper, Ollama, and NLP pipelines, and a Hockey Data Analysis app built with Flask, Streamlit, and Docker.