Denys Zhak
software engineer · distributed systems & ML infra
Most machine learning outputs are predictions someone reads. I work on the kind that moves physical equipment – at Imubit, I build the backend and platform around deep-learning controllers that run refineries in closed loop.
I spend most of my time on serving, monitoring, and keeping production ML systems reliable. Recently I've been contributing to ty, Ruff, Docker Model Runner, and a couple of GPU inference projects using Vulkan and ROCm. I write here about problems I've run into and what I learned solving them.
tl;dr
- 9 years building and operating data-intensive systems, the last five for ML
- Extensive experience in the Python ecosystem, 3+ years of production Go, and a recent focus on Rust
- open-source contributions to developer tools and AI infra – ty, Ruff, Docker Model Runner and others
- I studied Air Navigation at the National Aviation University in Kyiv
contact
I'm always happy to hear from people working on distributed systems, ML infra, dev tools, or inference: denyszhak@gmail.com.