I develop infrastructure where performance and resource use directly affect production systems: packet processing, load balancing, traffic protection, and distributed services.
At Ozon Tech, my work spans C/C++ networking systems, Go-based Kubernetes controllers, build and release tooling, and client-informed CDN traffic steering. Earlier, I worked on advertising infrastructure at VK and medical-image geometry at Align Technology.
How I got here
I came to GPU computing out of curiosity about the hardware in my laptop. In 2019, I independently built a C++/CUDA framework for wave simulation, then compared its output with theory and colleagues’ physical experiments.
My first commercial role was in iOS development. I later moved towards backend and infrastructure work because I wanted to solve problems where performance, resource costs, and business value meet. I enjoy understanding unfamiliar code and taking work from investigation and design through implementation and production rollout.
Systems work
At Ozon, that includes maintaining internal yanet and Envoy forks, improving network-policy update performance, and building the tooling needed to test and release C++ systems. I also designed a cross-cluster CronJob controller and work on collecting and applying client-side CDN measurements.
At VK, I worked on ad-serving identifiers, auction-object caching, an advertising gateway, and container-platform adaptation. Investigating memory use meant manually reconstructing a complex dependency graph and collaborating with neighbouring service owners. My changes supported a broader team effort to shard the platform.
Working across disciplines
At Align Technology, I developed geometric algorithms for combining medical-image representations and supporting orthodontic planning workflows. I also worked on release validation and risk reviews, and learned to explain technical choices to doctors and technicians in a multicultural team.
At Aktiv, I built test orchestration and platform-specific runners and worked with certification processes for cryptographic devices and libraries. This gave me experience with standards and the practical constraints of complex R&D work.
Engineering approach
I want to get the most from a system with the least unnecessary effort. I am interested in code close to the hardware, but the purpose of optimisation is practical: making useful systems run efficiently with the resources available.
Mentoring
In 2023, I mentored students on VK Education’s Distributed Systems in Go course.
Education
I am currently studying Informatics and Computer Engineering at MISIS.