At XYMA Analytics, I engineer production machine learning systems for refineries, aluminum and steel plants, and power generation. My models support early failure detection, corrosion and temperature prediction, pot health assessment, and yield optimization.
I own end-to-end ML pipelines using PyTorch, TensorFlow, Docker, and Kubernetes for scalable deployment and real-time inference. I also trained FoxSLM, a compact language model from scratch in PyTorch for industrial report generation on CPU-only edge hardware.

