I started in 2000 as a junior engineer at Sebha University, keeping campus servers
and systems running and building the internal applications the administration relied on.
A master's degree at the University of Malaya took me deep into networking and data
communication — raw TCP/UDP sockets, secure transfer, and a research project that turned
iris patterns into digital signatures backed by elliptic-curve cryptography.
That research thread led to a Ph.D. in Electrical & Computer Engineering at Toronto
Metropolitan University, where I spent years teaching neural networks to see hearts:
U-Net-based models that segment the left ventricle in poor-quality echocardiograms with
~0.97 Dice accuracy, fast enough (~0.05 seconds per image) for near real-time clinical use.
That work became first-author papers in peer-reviewed journals.
Most recently, at THEOSYM in Toronto, I built full-stack products with React and Flask and
designed APIs that orchestrate multiple LLMs — OpenAI, Claude, LLaMA, Gemini — selecting
models dynamically per request. The problems I keep coming back to are the same ones I
started with: system performance, real-time behavior, and carrying an idea from conceptual
design all the way to deployment.
