About

I earned my master's in Applied Mathematics from Lomonosov State University, where I worked on Extreme Value Theory in the Laboratory of Probability.¹ After graduating, I moved into the machine learning industry.

I am an ML engineer with a strong interest in applied research: understanding how models behave, testing ideas rigorously, and turning promising results into reliable products. My experience spans predictive modeling, distributed data processing, generative AI, recommendation systems, and MLOps.

I started at MaximaTelecom, building classification and regression models for tabular and text data, as well as large-scale Apache Spark pipelines. Later, as the fourth technical hire at a GenAI startup, I owned a core product feature end to end—from training and fine-tuning generative models in PyTorch on multiple GPUs with FSDP to production deployment.

More recently at Magical Labs, I have worked on research in latent diffusion steganography. I developed provably secure covert messaging method for embedding high-capacity watermarks that remains robust under aggressive JPEG compression. Along the way, I reproduced and implemented relevant research papers and open-sourced an implementation of one approach with results.

Alongside industry work, I co-authored and taught a machine learning course for 500+ students at PASV.