Computer vision · Machine learning · Systems
MASc researcher and AI engineer building production-oriented perception and explainable agentic systems - from GPS-denied aerial robotics to evidence-grounded research workflows.
About
I work where model quality meets operational reality. My focus is on perception and AI systems that make their reasoning inspectable, stay within real hardware constraints, and keep working when the environment gets less predictable.
That shows up in resource-efficient visual-inertial navigation, source-aware research agents, temporal vision research, and latency-sensitive platform software. Across each, I like turning a difficult technical seam into something measurable, reliable, and easier for the next person to understand.
Experience
University of Waterloo / TensorOne Inc.
TTEC
VIP Research Group, University of Waterloo
General Motors
Western University
Genpact
Selected projects
A selection of systems that show the range: edge perception, careful emulation, and real-time graphics.
A production-oriented visual-inertial system that combines CPU VIO, map matching, and quantized self-supervised learning on Raspberry Pi 4.
A deterministic C++ emulator modeling the CPU, memory, DMA, timers, interrupts, and LCD/PPU state machines at the cycle level.
An interactive scene with dynamic lighting, Gerstner waves, tessellation shaders, and a later Metal port from OpenGL/GLSL.
Education
University of Waterloo
Research focus: machine learning. Thesis focus: resource-efficient VIO/SLAM and online sensor recalibration for aerial robotics.
Western University
Skills
Let's talk
I am always glad to connect about computer vision, machine learning, systems engineering, and research collaboration.