Artyom Arutyunyan

About me

Personal

Christian, husband, father. Hiking and bushcraft enthusiast. Film photographer. OSS contributor and maintainer.

Education

2010 – 2015, Synergy University, specialization in Information Security.

English: intermediate (B2).

Rates

Full-time: open to offers from $120,000/year.

Part-time: $100/hour — ready to help you with refactoring or performance.

AI-Augmented Engineering

I use AI coding agents as force multipliers for implementation and review, while keeping ownership of architecture, quality gates, and product decisions.

cursor-agent

What I bring to the table

LLM Landscape

I work hands-on with leading closed models (OpenAI, Anthropic, Google, Grok) and open models (DeepSeek, Qwen, Kimi, GLM), tracking how both sides of the frontier evolve.
I test smaller local models (Qwen, Mistral, Nemotron, Gemma) to understand where they are practical, where they fail, and what trade-offs they bring.

Building AI-Powered Features

I built a chat app from scratch, including streaming responses, tool calls, and model switching. That work taught me where LLM integrations get messy: every provider has its own request format, streaming protocol, error model, and rate-limit behavior. To keep the app maintainable, I used an AI toolkit and OpenRouter as a unified layer, then extended the system with MCP servers that connect the app to external data and actions.

Prompt Engineering & Evaluation

I structure prompts across company, project, system, and task levels so agents have clear goals, plans, and constraints.
I design context strategies with skills, RAG, subagents, and task-specific memory instead of relying on one oversized prompt.
I build quality loops around agent work: harness hooks, automations, regression checks, and cross-model review.

Team AI Adoption

I helped establish AI engineering guidelines for a large engineering organization.
I defined practical workflows: which tasks to delegate to AI, where review gates belong, and what quality checks are required.
I helped design a verification pipeline for agent work, combining automated tests with stronger human review.
I mentored engineers on prompting, harness setup, and choosing the right level of context.
I helped set security boundaries around cloud APIs, local models, and sensitive data.
I measured productivity impact and used the results to refine team practices.

Harnesses & Developer Tools

I have compared and used AI coding harnesses including Claude Code, Opencode, Kilo, and Cursor.
I work hands-on with MCP, both by integrating existing servers and by building custom servers for product tooling.
I am building a GitLab review-agent bot to bring AI-assisted code review into a controlled engineering workflow.

Experience across ambitious teams

Engineering leader with hands-on delivery across fintech, infra, and product organizations: I supported a banking app with 30K concurrent users, operated a central bare-metal datacenter and 1,000+ remote Cisco/server nodes for a national electronics retailer, launched a mobile banking product as Product Owner, built and led a full-stack gamification platform for enterprise support, and worked as a developer in 3 of the country’s top 4 banks. Alongside extensive freelance frontend work, over the last 7 years I have focused on React, TypeScript, and Node.js — authoring libraries, solving performance and architecture bottlenecks, and leading engineers technically while driving DevRel initiatives.

Frontend Developer (Lead)

Cloud.ru

Jun 2025 — Present

Details

Maintain core billing systems
Proactively identified and resolved a critical performance issue for a key enterprise customer
Built a documentation service for internal developer documentation
Delivered talks and conducted reviews at internal meetups
Helped integrate AI tools and establish engineering guidelines for all developers
Earned three achievement awards in my first half-year and ranked in the top 150 employees company-wide (out of 2,000)

Frontend Developer (Senior)

SimBaze

Sep 2021 — May 2025

Details

Refactored the entire network layer of a 500K LoC product to improve stability and performance
Refactored a large, unreliable routing system by introducing type-safe routing; reduced route-related regressions from ~1/month to near zero and improved refactoring confidence across the team
Found and fixed a security vulnerability in a third-party library
Integrated Gitpod workflows to reduce infrastructure costs per developer
Brought in a few key developers through networking

Fullstack Developer (Lead)

Kaluga Astral

Jun 2020 — Sep 2021

Details

Architected an internal gamification service from the ground up, driving full-stack design and delivery
Led an internal hackathon to ship a critical feature under tight deadlines — a real-time collaborative schedule editor supporting 20+ concurrent users
Built a lightweight internal framework that standardized Telegram bot development across the team
Introduced internal low-code tools: one to streamline backend workflows for engineers and a visual CMS for non-technical managers, enabling self-service and freeing engineering time
Mentored engineers and drove internal DevRel initiatives

Freelance Frontend Developer

Independent

2017 — 2020

Details

Delivered client web applications from discovery to release
Built responsive UI systems and reusable component libraries
Coordinated requirements, timelines, and handoffs with clients

JavaScript Developer

CSSSR

Jan 2018 — Aug 2019

Details

Built client-side web applications and clarified business requirements
Owned architecture choices across the stack, domain modeling, and tooling
Established code quality standards for docs, linters, tests, and tech debt management
Managed team allocation, training, and development materials
Delivered EDMS projects and produced talks and articles