I’m interested in machine learning, AI, and scientific computing, especially projects where theory meets practical implementation.
Most of what you’ll find here are learning-driven and research-style projects: building models from scratch, experimenting with neural architectures, and creating end-to-end pipelines for data, training, and evaluation.
-
ML library from scratch
Traditional ML models and neural network layers implemented using low-level numerical operations to better understand optimization and model behavior. -
Character-level Transformer
A decoder-only Transformer trained to generate Harry Potter–style text. -
Active learning for molecular potentials
Scripts and workflows for training machine learning potentials with active learning. -
Small side projects
Including a simple Python-based Tetris game.
- Python, C, SQL, R, Julia, MATLAB, Shell, Scala
- NumPy, Pandas, scikit-learn, PyTorch, TensorFlow
- Git, Linux