Ahmed ElSaid

I'm an Assistant Professor of Computer Science at the University of North Carolina Wilmington, where I work on the Intelligent Systems Engineering program.

My research builds machine learning systems for real-world engineering and scientific problems — the kind where data is noisy, multimodal, and expensive to get wrong. I work primarily with metaheuristic and nature-inspired optimization — search methods for spaces too large or irregular for gradient-based approaches. I apply these to automating the design of neural networks, and to optimization problems beyond machine learning, including on quantum architectures.

The applications span aviation, energy, agriculture, marine science, and climate: predicting engine failures from flight recorder data, forecasting power plant behavior, detecting crop-damaging bird flocks from drone imagery, and forecasting seasonal rainfall from ocean conditions. I also work on the systems side — embedded sensing, edge computing, and high-performance computing — since deployment constraints usually decide what models are actually viable.

Before my graduate school, I spent several years as an aircraft maintenance and airworthiness engineer. That's where most of these problems came from, and it's still how I choose them.