Hi, I’m Ilja Baroŭski, a Master of Mathematics and Computer Science. I graduated from the Faculty of Mechanics and Mathematics at Belarusian State University in 2023 with a Bachelor’s degree in Mathematics and System Analysis. In 2025, I earned my Master’s degree in Mathematics and Computer Science.
I have experience with a variety of programming languages,
including C/C++, Rust, Python, Julia, C#, Dart, Wolfram
Language, Go, and Crystal. My approach to programming is rooted
in improving efficiency, optimizing algorithms, and building
tools from the ground up. You can explore some of my projects on
the projects
page.
For my master’s thesis, I focused on computer modeling of the human voice. The goal was to simplify the model using wavelet theory, making it lightweight yet effective. As part of this work, I developed two key libraries:
- waveletnn: A library for integrating wavelet transforms into PyTorch models, enabling efficient neural network architectures for signal processing.
- ubpe A Universal Byte-Pair Encoding library designed for general sequences, including text and quantized signals. UBPE stands out for its ability to generate multiple encoding variants and its optimized implementation for diverse data types. For a deeper dive into UBPE, check out my detailed articles on its design and implementation: training process, encoding and decoding.
Beyond technical work, I’m continuously exploring ways to bridge theoretical research with practical applications. Whether it’s through open-source contributions, writing, or collaboration, I aim to create tools that are both innovative and accessible.