A summary of my work, research, and learning experiences across AI, graphics, and software engineering.
Developed and maintained the company's glare-control rendering application while conducting AI research into brightness interpolation and perceptual consistency. Implemented novel image interpolation techniques inspired by state-of-the-art papers in generative modeling. Worked extensively with Rust and Python, and trained models on Linux-based HPC infrastructure.
Recommendation letter available upon request.
Created custom 3D rendering, raytracing, and visualization applications for clients using C++, OpenGL, and GLFW. Delivered optimized and visually rich projects, often involving real-time graphics and interactive design.
Collaborated with the owner of the popular League of Legends Pictionary website to integrate an AI-based character classifier trained on user-submitted sketches. Designed and deployed the data collection pipeline, preprocessing scripts, and trained models, which were permanently integrated into the production site as a new game mode.
Authored and published a research paper analyzing doodle classification methods using RGB-domain sketch data. Compared CNN and transformer-based models, visualized feature embeddings using t-SNE, and demonstrated fine-tuned ResNet-50 marginally outperformed human accuracy.
Read on arXivDemonstrated proficiency in building and deploying deep learning models with TensorFlow, covering computer vision, NLP, and data preprocessing.
View Certificate