I get a kick out of turning research ideas into fast, testable code — usually with Rust or C/C++ for performance and Python for experimentation/flexibility.
- Reinforcement- & deep-learning algorithms — from classic q-learning/policy-gradients to encoder/transformer-based processes
- Distributed RL frameworks for HPC and robotics
- Experimental OS research ranging from the kernel to interactions with userland
- Neurosymbolic SAT engines that blend logic with learning
- Generative UX tools — e.g., GAN-driven mouse-trajectory synthesis, adaptive song equalizer pipeline, etc.
- Computer-vision apps that leverage CV models for real-time use
LinkedIn • open to collaboration and feedback