MSK · UTC+3 · 18 Nov 2026one engagement open · Dec–Jan

About

Ilya Rutj · Moscow, UTC+3

I have been doing systems work since 2013: first as the only admin of a small hosting company, then for teams that had grown faster than their infrastructure.

How I got here

My first paid job was rebuilding a mail server that had been running on a machine nobody could log into. The previous admin had left a note with three passwords and no explanations. That week taught me the thing I still optimise for: infrastructure that a stranger can understand at 3 a.m.

Since then I have worked with about forty setups — hosting panels, a regional ISP edge, a two-person SaaS, a media site that got mentioned on television and fell over in forty seconds. Most of the work is unglamorous: backups, renewals, monitoring, and deleting things.

How I work

Engagements
2–6 weeks, one at a time. I would rather finish one properly than juggle three.
First week
Read-only. I inventory what exists, write down what I do not understand, and send you the list before changing anything.
Changes
One at a time, with a rollback that I have tested. No silent Friday deploys.
Handover
A written runbook in your repository, in the language your team speaks, plus one call where I read it to you.
What I don't do
Frontend, mobile apps, and anything that requires me to be on call for a system I did not build.

Tools I actually use

Debian and Ubuntu, nftables, systemd, nginx, Xray/Reality for tunnels, restic for backups, Prometheus and a small Grafana, Zabbix where it already exists. tcpdump and ss more than any dashboard. A text editor and a notebook.

Elsewhere

I answer email within a day: hello@ge4.externcashpn.cv. I do not have a newsletter, and I am not on LinkedIn. Notes are written when something breaks and I learn why.