Welcome to my portfolio! I'm a researcher working towards my Master's in Computer Science at KAUST. I love tackling tough and intellectually stimulating problems across a range of CS topics. Scroll through this page to learn more about my interests and some of my proudest work.
My current interests surround parallel and high-performance computing, and how our algorithms and data-structures can take full advantage of them.
During my BS at Boston University I learnt how to design and fine-tune neural network pipelines for tasks ranging from chess and backgammon to geological thin-sections and cancer classification.
My Master's has focused on using parallel computation and intelligent algorithms to optimize the execution of heavy-duty computations like visual ray-tracing, image compression and acoustic beam-tracing.
Scientific models and simulations provide some of the most important (and satisfying) applications of HPC & AI. Much of my work in those fields intersects with topics like acoustics & computational fluid dynamics.
A mix of my most challenging and rewarding projects through the years.
(MSc Thesis) Multi-GPU implementation of Gaussian-Beam Tracing for acoustic simulation. Handles arbitrary user-defined meshes and noise-sources, and visualizes and plots results.
Python C++ CUDA CuPy PyVista
Raytracing simulation built with no external libraries in C++, which handles shading, shadows, reflection and movement & rotation of objects.
C++
SVD and quantization based compression of the large-scale Red Sea temperature dataset (shown above).
Python PyTorch
GPU-powered heat transfer simulation using CUDA and finite differences.
C++ CUDA Python
(BS Capstone) Adversarial architecture which trains a chess agent by penalizing non-humanlike moves.
Python PyTorch Pandas
Backgammon playing agent which uses probabilistic Monte-Carlo tree search for decision-making.
Python Tensorflow