I write production code through my own AI-agent delivery loop — backend, infrastructure, security.
_{Open to remote work, UTC+3}

• 💫 About Me
• 🚀 Areas of Interest
• 🧰 Technology Stack
• 🤖 AI & Data
• 🔐 Security Work
• 🧭 How I Ship
• 📫 Contact

Backend Developer , I create high-performance APIs and automated solutions for Telegram and VK.

In my free time, I dive into the world of cybersecurity: analyzing malware, researching vulnerabilities, doing OSINT, and configuring secure infrastructure.

Delivered engagements

• Web application penetration test — 46 findings, 7 of them P1; report handed over.
• Network measurement research — a 21-page paper backed by 66 sources on traffic classification and censorship behaviour.

Tooling

Every code change takes the same path — no hand-edited builds on the server.

flowchart LR
    A[Change] --> B[Quality gate<br/>lint · types · tests]
    B --> C[Image<br/>distroless build]
    C --> D[Deploy<br/>compose · Ansible]
    D --> E[Verify<br/>routes · logs · DB]
    E -->|regression| A

• Change — one topic per branch, conventional commits, review before merge.
• Quality gate — linters, type checks and tests must be green; a failing gate blocks the merge.
• Image — distroless, non-root, reproducible builds.
• Deploy — Docker Compose or Ansible across nodes; secrets are not baked into images.
• Verify — probe the deployed artifact: real routes, real logs, real rows.