Yuri Trukhin.

trukhin.com · platform products · est. 2007

Cloud & AI Platforms Leader

Product, engineering and economics of platforms that people build on.

I’m Yuri Trukhin (Iurii Trukhin, trukhinyuri on GitHub). At MTS Cloud I ran a public-cloud development center — 139 positions in the staffing plan — and built its GPU service line on NVIDIA V100/vGPU. At JetBrains I worked on Developer Experience: WSL, Docker, Dev Containers, Terminal. Now, at CloudLinux and remote from the Netherlands, I work on infrastructure reliability, FinOps decisions and agent tooling. Every claim on this site carries its evidence boundary.

The full story GitHub LinkedIn [email protected]

Proof, with boundaries

Three cases I can document. Each states the role, the scope, the outcome — and the boundary of what it proves. I would rather under-claim than decorate.

MTS Cloud · 2018–2023

Cloud development center & GPU product line

Role
Product manager → Head of Development Department → Head of Development Center (from October 2021)
Scope
Public cloud platform; development center of 139 positions (128 filled) in the staffing plan
Outcome
Built the center’s engineering organization; launched the GPU cloud service — NVIDIA V100/vGPU configurations, tariffs, client training and inference scenarios (2019–2020)

Boundary: 139 positions is the center’s size in the staffing plan, not the number of direct reports. Client-effect forecasts from that period are not claimed as results.

Public record: AI Journey 2020 talk →

JetBrains · 2023–2025

Developer Experience for IDEs

Role
Developer Experience: WSL, Docker, Dev Containers, Terminal, remote environments
Scope
Developer-tooling surface used across JetBrains IDEs
Outcome
WSL integration research with a named owner; joint work on remote development environments; Dev Containers and Kubernetes debugging demos shipped publicly

Boundary: team products — sole authorship is not claimed; adoption and NPS figures from that period are not verified and not cited.

Kubernetes debugging demo →

CloudLinux · 2025–now

Reliability, FinOps, agent infrastructure

Role
Platform infrastructure: reliability and infrastructure decisions, FinOps, agent tooling
Scope
US-based OS and platform company; remote from the Netherlands
Outcome
PostgreSQL migration: data-protection requirements set, plan corrected after review, change accepted with the consumer confirming database stability. FinOps: an AWS commitment case built on a line-by-line reconciliation with Finance, with coverage checks and a strict projection-vs-measured separation

Boundary: the migration itself was executed by an engineer — my role was directing the change and acceptance. In FinOps, process adoption is proven; realized savings are not claimed.

Agent infrastructure work →

Now

What I am actually working on, dated. Updated when the work changes, not on a schedule.

  • Agent infrastructure, in the open: maintaining codex-superpower (a fork of a multi-model agent toolchain — ~30 ported fixes, each with a regression test, self-update with rollback) and CloudRING (public Go core, Apache-2.0). agent-tools — a single read-only gateway for agents to external systems — stays private; the pattern is public.

  • This site, rebuilt as v2: pure static HTML, one hand-written stylesheet, zero JavaScript, zero external requests — built to be fast for people and legible for AI agents (llms.txt, JSON-LD, sitemap). First writing post: GPU platform economics.