Open to new projects

I ship AI products that actually sell

I ship full-stack AI products: polished interfaces, APIs, and LLM systems that hold under load. Fast, reliable delivery that grows your product.

  • Fixed estimates, no hidden fees
  • Weekly demos in sprints
  • 24h response time
  • Source code & IP yours
Languages TypeScript · Python · Go
Frontend React · Next.js
Backend Node.js · PostgreSQL · Redis
Infra Docker · Kubernetes · AWS · CI/CD
AI & data LLM · RAG · LangChain · Vector DBs
  • 8+years of experience
  • 150+projects
  • 40+clients
  • 98%happy clients
developer.js
const developer = {
  name: 'Alex',
  role: 'Full-stack Developer',
  stack: ['React', 'Node.js', 'Go'],
  ai: ['LLM', 'RAG', 'Agents'],
  focus: 'AI-first products & UX',
  remote: true,
  openToWork: true
};
LLM React Agents
FuryCow mascot - a cow developer coding at a laptop

// 01 · about

About me

I'm a programmer who loves writing clean code and building understandable architectures. Over 8 years I've gone from slicing landing pages to designing distributed systems that handle hundreds of thousands of requests per minute.

I work with product teams end to end: from problem research and prototyping to release, monitoring and support. I believe good engineering is not just code — it's honest deadlines, measurable results and comfortable communication.

  • LocationMoscow · remote worldwide
  • SpecializationFull-stack · Web · SaaS · AI/LLM
  • LanguagesRussian · English (B2)
  • Currently learningRust and edge computing
0
years of experience
0
projects
0
clients
0
satisfaction

// 02 · services

How I can help

From idea and prototype to production and support

AI & LLM engineering

LLM-powered features, RAG pipelines, AI agents and tool loops. From prompt design and evaluation to production — cost-efficient, reliable, without hallucination surprises.

DevOps & Cloud

CI/CD, Docker, Kubernetes and cloud infrastructure. Reliable deploys, observability and cost-aware setups that stay maintainable as the product grows.

Bots & integrations

Telegram bots, webhooks and third-party integrations. Automate workflows, notifications and CRM glue so teams spend less time on manual ops.

Architecture & consulting

Technical audits, system design and MVP scoping. Clear trade-offs, a realistic roadmap, and hands-on help from prototype to production.

Web applications

SPA and SSR with React / Next.js: dashboards, marketplaces, customer portals. Clean UI, accessibility and fast load times.

Backend & API

Node.js and Go, REST and gRPC, database schema design, integrations with payment and external services.

// 03 · projects

My projects

Selected work from recent years

CodeCritic — AI code review for GitHub PRs Live

CodeCritic — AI code review

A SaaS that reviews GitHub pull requests with AI: in the browser, as a GitHub Action and via REST API. Free tier, team features and an option to plug a custom LLM.

  • Automated review of every PR — before a human ever looks at it
  • Three integration paths: web UI, GitHub Action, API
Ruby Grape API PostgreSQL React LLM
MentionVox — AI search visibility for crypto and Web3 Live

MentionVox — AI search visibility

A GEO platform for crypto & Web3 brands: scores crawler health, structured data and mention likelihood in AI search versus competitors, with a free snapshot in about a minute.

  • Brand visibility scoring against competitors in ~1 minute
  • Multilingual product: English, German and more
Next.js TypeScript GEO / SEO LLM
memory_graph_mcp — MCP server: code knowledge graph Open Source

memory_graph_mcp — code knowledge graph

An MCP server that gives coding agents a dependency graph of the codebase: query context for a file or symbol, run impact analysis before a change, and surface implicit side effects like DB writes or external calls.

  • Impact analysis before editing: affected nodes and risks
  • Works with any MCP-compatible agent (Claude, Cursor, etc.)
MCP TypeScript AI Agents Static Analysis
openrouter_agent - agentic IDE for the OpenRouter API Open Source

openrouter_agent - agentic IDE

An agentic IDE focused on the OpenRouter (OpenAI-compatible) API: agentic models with tool loops - read files, patch code, run tests - routed across models from one key.

  • Tool loops out of the box: read, edit, run, verify
  • One API key for every model on OpenRouter
TypeScript AI Agents OpenRouter LLM

// 04 · writing

Latest writing

Notes on full-stack & AI/LLM engineering

All posts →

// 05 · experience

Work experience

Key milestones from recent years

2024 — present

Team Lead · FinTech

Leading development of payment products end to end: a team of 6 developers, weekly releases, on-call rotation.

  • Set up AI-assisted code review on every pull request: bugs caught before merge, incidents down 45%
  • Rewrote the reconciliation service: API response time dropped from 1.2s to 300ms
  • Mentoring 3 senior-level developers and 1 senior QA engineer
React · Node.js · PostgreSQL · LLM
2020 — 2023

Senior Full-stack Developer · FinTech

Built microservices for a payment bus serving fintech products with strict reliability requirements.

  • Designed the payment service: 2M transactions per day, 99.95% SLA
  • Migrated a monolith to event-driven architecture (Kafka)
React · Go · Kafka · PostgreSQL
2018 — 2020

Frontend Developer · PixelWorks

Built interfaces for client projects: e-commerce stores, corporate portals, landing pages.

  • Created a design system that doubled the speed of building typical pages
  • Optimized loading: mobile performance improved 2.5x
JavaScript · Vue · SCSS · Webpack
2017 — 2018

Junior Developer · freelance

First commercial projects: WordPress websites and early SPAs. Started maintaining open-source pet projects around the same time.

HTML/CSS · JavaScript · WordPress

// 06 · skills

Skills

Technologies I work with every day

AI / ML / LLM

LLM integration 90%
RAG pipelines 85%
AI agents & loops 80%
Prompt engineering 90%
ML fundamentals 70%

Frontend

JavaScript 90%
TypeScript 85%
React / Next.js 85%

Backend

Node.js 80%
Python 75%
Go 70%

Data & infrastructure

PostgreSQL 80%
Docker / Kubernetes 75%
CI/CD 70%

Practices

Git / Code Review 90%
Testing 75%
Architecture 80%
Also working with: Redis MongoDB GraphQL Kafka OpenAI Anthropic LangChain Vector DB PyTorch AWS Terraform

// 07 · reviews

Client reviews

What the people I've worked with say

"In two months Alex got our product in order: rewrote the slow part of the API and set up CI/CD. For the first time in a year, releases went out without incidents."
Emily CarterNDA
"A rare quality — he understands the business, not just the code. He proposed a solution that saved us a quarter of development time and a lot of stress."
James WilsonNDA
"We worked remotely across three time zones — communication was better than in an office. Always a clear status, risks and next steps."
Anna LangeNDA

// 08 · faq

FAQ

What does the process look like?

We start with a short call to break down the task and goals. Then you get an estimate with stages and timelines, followed by development in 1–2 week sprints with a demo at the end of each. You always see the status in the tracker and the current build.

How much does a project cost?

It depends on scope: a small MVP takes a few weeks of work, a full platform starts at a couple of months. After the brief I provide a fixed estimate per stage, with no hidden fees.

Do you work with legacy code?

Yes. I start with an audit: finding bottlenecks, risks and quick wins. Then — incremental refactoring without stopping the business, with tests covering the critical logic.

Do you support the project after release?

Yes, there's a technical partnership format: monitoring, dependency updates, backlog improvements and an SLA for response time. Alternatively, I hand the project over to your team with full documentation.

How quickly do you respond to messages?

During working hours — within a couple of hours, 24 hours at most. Communication agreements are fixed at the start: preferred channel, call frequency and time zone.

// 09 · contact

Get in touch

Ready for a new project or just want to discuss an idea? Tell me about your task — I'll get back with an estimate within 24 hours.

↑