Kirill M

Automation Engineer

Apr 2019 - Present • 7 yrs 6 mos

Python automation and AI/LLM integrations for small businesses: autonomous content pipelines, Telegram and Discord bots, API-to-API integrations, web scraping and data extraction, self-hosted services with monitoring and alerting. The system I use as my own reference has been publishing continuously since 2016 and has been rewritten from scratch four times; the version running now went live in May 2026. It needs no manual work: it ingests sources, scrapes each item, and an LLM writes a summary, tags and an image before it publishes - about 1,150 items a month, and that is a setting rather than a ceiling: on free provider tiers alone the same pipeline has headroom for up to ~50,000. It runs on a cascade of roughly ten LLM providers. Every stage has its own fallback chain, providers validate their own credentials at startup and drop out silently when a key is revoked, and two stages have a fallback for when the fallback chain is already empty. The end state is always degraded output, never a stopped pipeline. Alerting is three layers, each one still working if the layer above it is dead: errors to a dedicated channel, an alert state machine that closes a failure automatically when the matching success arrives, and a systemd hook that needs neither the app nor the database to reach me. That is the difference between an automation you trust and one you check on every morning.

Senior Systems Administrator

Apr 2020 - Jan 2023 • 2 yrs 9 mos

Hopper Support LLC. Ran production infrastructure for large-scale, nationwide client projects under 24/7 SLAs - CentOS and RHEL, nginx, PostgreSQL and MySQL, Docker, GitLab, Zabbix, Redis, Let's Encrypt. Designed and maintained the architecture for client web applications, including Java services and LEMP stacks with Node and Redis. Built GitLab CI/CD with automatic deployment of QA test stands on a Traefik + GitLab + Ansible pipeline. Migrated a 50+ GB Jira and Confluence installation off Atlassian Cloud onto self-hosted servers and then into a data centre, on a hard deadline and without losing history. Most of this job was finding the things that had quietly been done the wrong way, before they became an outage.

Systems Architect

Apr 2019 - Nov 2022 • 3 yrs 7 mos

Atlant - sole IT owner for a small company, building everything from scratch: servers, monitoring, backups, and the day-to-day automation of internal workflows. Troubleshooting, organising workflows, automating repetitive tasks, and rolling out new capabilities as the business needed them. Small-company IT is where you learn that the correct solution is the one the client can live with, not the one that looks best on a diagram. There was no second engineer to call.