About

Summary

Backend engineer with 10+ years building high-load systems across Java, PHP, and Node.js. Architected a microservice platform from scratch serving 1M+ daily unique users and handling 40M+ requests/day — up to 1K RPS at peak. Lead a team of eight engineers while staying hands-on in the codebase.

Known for pragmatic, unconventional solutions — from agentic AI workflows to a self-hosted GPU rig running local ML workloads under the desk.

Skills

Languages

Java 17–21+, SQL, Node.js, PHP.

Java ecosystem

Spring Boot 3, Spring WebFlux, Spring Security, Spring Cloud, Hibernate, R2DBC.

Data & messaging

PostgreSQL, MySQL / MariaDB, Redis, ElasticSearch, RabbitMQ.

Infrastructure & DevOps

Docker, Docker Swarm, Kubernetes, Ansible, Nginx, Linux (Ubuntu / Debian / Alpine), Bash.

CI / CD

GitLab CI, GitHub Actions, Bitbucket Pipelines.

Frontend & mobile (AI-augmented)

Vue.js, React, React Native, Tailwind CSS, Flutter.

AI-assisted development

Claude Code, Codex, GitHub Copilot — agentic coding workflows, multi-agent setups, AI teammates for code generation, reviews, and architectural prototyping.

Self-hosted AI / ML infrastructure

NVIDIA GB10 Blackwell, Ollama (Qwen, Gemma 3), Whisper, ComfyUI, n8n automation pipelines. See the desk.

Experience

Java Team Lead — Project Vanta · 2021 → now

High-load ad-tech platform on a microservice architecture. Lead backend engineer for a cross-functional team. Confidential (NDA).

• Co-architected the platform alongside founding engineers and led the team that took it to ~1M daily unique users, 40M+ requests per day and peaks around 1K RPS, holding sub-300ms p95 on the hot paths with 99.95% availability.
• Grew backend from a solo contributor into a two-tier team of seven — hiring, onboarding, pairing — and mentored one engineer into a sub-lead role with full ownership over a vertical.
• Shifted incident handling from reactive firefighting to structured on-call: SLO-based alerting, weekly incident reviews, and runbooks. MTTR dropped by roughly half in the first quarter after rollout.
• Partnered with product, QA, and business on release planning. Client-facing regressions now ship fixes in the same business day instead of the next sprint.
• Introduced agentic AI tooling (Claude Code, Codex) into the team's workflow — measurable lift on shipping speed and review-time defect catch rate. AI is also used end-to-end to deliver fullstack and mobile features.

Stack: Java 21, Spring Boot, Docker, Ubuntu.

Backend developer & DevOps — B2B software studio · 2016 → 2021

Multiple client projects inside an outsourcing company. Based in Russia until June 2019, then relocated to the EU.

• Project Lot — online real-estate listing platform with CRM. Backend microservices for search and CRM workflows, GitHub Actions CI/CD, containerized Nginx / Docker infra on Linux.
• Project Beacon — influencer marketing & lead-generation platform. Backend features and critical production bug triage across backend and Android. Sprint planning with the dev team.
• Project Hearth — messenger for emotional support with bot chats and groups. REST API and SPA admin from the ground up; real-time messaging with RabbitMQ + Redis; CI/CD via Bitbucket Pipelines to AWS with Docker Swarm.
• Project Tempo and others — fitness platform, booking systems, marketing chatbots. Backend and frontend features across many smaller client projects.

Stack: PHP, Laravel, Lumen, Java, Node.js, Vue.js, React, Angular, PostgreSQL, MySQL, Redis, RabbitMQ, ElasticSearch, Docker, Docker Swarm, Nginx, AWS, Azure, GitHub Actions, Bitbucket Pipelines.

Earlier career · 2015 → 2016

PHP developer at a web studio, then a classified-ads startup. First production experience with PHP, MySQL, and frontend basics.

Education

Computer Systems Software — Technical College at East Siberian State Technological University. 2009 → 2013.