Summary
Timothy Castiglia is a Machine Learning researcher and Field Consultant with a PhD from Rensselaer Polytechnic Institute and a decade of experience building efficient, privacy-aware AI systems. He specializes in federated and distributed learning, having developed novel feature-selection and compression techniques for vertical federated learning at IBM that led to patents and ICML publications. Comfortable in both research and production settings, he pairs deep ML theory with systems work—implementing experiments in PyTorch/TensorFlow and contributing Rust, C/C++, and Python backend code for real-time and embedded applications. Timothy focuses on reducing the time, monetary, and ecological costs of large-scale AI while tackling privacy, convergence, and communication constraints across globally distributed participants. He is now applying his broad skill set to practical problems that make AI safer, cheaper, and more accessible to everyone.
9 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, 4.0/4.0, Doctor of Philosophy - PhD, 4.0/4.0 at Rensselaer Polytechnic Institute