Remote for product teams worldwide — from architecture to post-launch support
Node.js, TypeScript, React/Next.js, Vue, PostgreSQL, K3s, and AI integrations. Bridging classical web stacks with modern AI workflows — from RAG pipelines to local inference.
Stack I work with
TypeScript
Node.js
Vue.js
React
Fastify
PostgreSQL
Go
Redis
RabbitMQ
Docker
Kubernetes
Linux
Ollama
Anthropic
LangChain
Grafana
Next.js
Nginx
Git
Yandex.Metrika
Google Analytics
Custom web applications built with modern frameworks. I handle the full stack — from responsive frontend to scalable backend APIs.
Scope
Next.js / React frontend with SSR & SEO
Node.js / NestJS backend APIs
PostgreSQL database design & migrations
Authentication, payments, admin panels
Responsive design for all devices
Fast loading on all devices
Deep analysis of your codebase with a prioritized action plan. The output isn't just a list of issues — it's a clear picture of what to fix first and which technical debt can safely wait.
Scope
Codebase health report (security, performance, maintainability)
Architecture diagram & improvement plan
Dependency audit & upgrade strategy
Database schema review & normalization
API design review (REST / GraphQL)
Technical debt prioritization matrix
I help distributed teams adopt engineering best practices through hands-on mentoring — not lectures. Fully remote across timezones: code reviews, pair programming, and building a culture of quality from within.
Scope
Weekly code review sessions
Team onboarding playbook
Git workflow & branching strategy
Testing strategy (unit, integration, E2E)
Code style guide & linting setup
Technical interview process design
Production-grade infrastructure that won't wake you up at night. From a single VPS under an MVP to multi-region cloud under load — set up so the team ships every day and sleeps on weekends.
Scope
Docker containerization & compose setup
CI/CD pipelines (GitHub Actions / GitLab CI)
Hetzner / DigitalOcean / Vultr / AWS provisioning
Nginx reverse proxy & SSL setup
Monitoring & alerting (Grafana, Uptime)
Zero-downtime deployment strategy
Bringing AI into products — from classic OpenAI/Anthropic APIs to local LLMs running on your own infrastructure. RAG systems, embeddings, vector indexes, and the economics of running models locally vs paying monthly cloud subscriptions.
Scope
Integration with Anthropic / OpenAI / Ollama / local models
RAG pipelines with LangChain orchestration and vector DBs (ChromaDB, Qdrant, pgvector)
Prompt engineering with Promptfoo evaluation harnesses
Embedding strategy with Cohere Rerank and indexing for large corpora
Local-inference infrastructure on owned RTX 3090 home server (Ollama, llama.cpp, GPU sizing)
Cost analysis: local hardware vs monthly cloud subscriptions
Non-standard requests that don't fit predefined services. One-off audits, technical consulting calls, narrow utilities or integrations, legacy-code reviews — discussed and scoped individually.
Scope
Initial scoping call (free, 30 minutes)
Individual scope and timeline tailored to the request
NDA on request before any code or context is shared
Hourly or fixed-price model — agreed upfront
All artifacts (code, docs, diagrams) belong to you
Direct Telegram communication during the engagement
Solo automation engineer
GitHub Actions · Fastlane · custom Bash scripts
Built the release pipeline for a multi-tenant mobile platform. 50+ Android and 50+ iOS apps shipping with a single command. Saved roughly one developer-day per release.
Founder, solo developer & maintainer
Node.js · React · PostgreSQL · Docker · native tests, no third-party libraries
My own product: a foodservice automation platform built from scratch. Multi-tenant architecture with data isolation between venues; menu, inventory, and finances in a single system. Live in production for several years — I keep developing and maintaining it. Full cycle: idea → architecture → deployment → operation.
Senior fullstack · consulting & development
Vue · Node.js · Go · K3s
End-to-end platform development: payments and payment-gateway integrations, metrics collection, performance optimization, internal system utilities. Large European-tier platform; client name not disclosed.
Independent R&D
Local LLMs (Ollama and others) · embeddings · RAG frameworks · ComfyUI
Researching local language models and building RAG systems without dependence on cloud APIs — including RAG for development workflows. Tracking the economics of local inference: electricity, input/output token cost, hardware vs monthly cloud subscriptions.
1-on-1 format — not lectures
JS/TS · React · Node.js · Git · SQL
Working with students from zero to their first job. Task-driven work and code review of real solutions, pair programming — practice over tutorials. Help with CVs and interview prep. Several students have successfully entered the IT industry and continue to grow.
Independent / DIY
Arduino · Node.js · Bash · system utilities · GitHub OSS
Personal projects outside paid work. Built an Arduino-based motor controller for home-automation actuators — retrofit push-button automation. Write small system utilities: for instance, a tool that reads temperature metrics directly from memory, available as an open repository on GitHub. Several utilities run on my own infrastructure.
An MVP usually takes 6–10 weeks. A production-grade app with auth, payments, and admin panel — 3–4 months. I always start with a detailed scope and timeline before writing any code.
Yes — a free 30-minute call before we start. We cover the task, context, timeline, and budget. If it's a good fit, I'll send a proposal within a couple of days covering scope, timeline, and price.
Both. About half of my projects start with an existing codebase that needs refactoring, performance fixes, or new features. I start every engagement with a code audit to understand what we're working with.
I start with a read-only audit: map modules, identify hot spots, and grade technical debt. Then I propose an incremental refactor — no 'rewrite from scratch' unless truly needed. Improvements ship as safe, reviewable pull requests so the product keeps running.
Yes, I sign NDAs before any code or internal docs are shared. All source code, designs, and specs belong to you from day one — work happens in a private GitHub/GitLab repo under your account or organization.
You do, always. Code lives in your repositories, on your infrastructure, from the first commit. If we pause or part ways, you keep everything: docs, deploy scripts, credentials. No vendor lock-in.
Sometimes, yes. Before committing I map the critical path and risks. If the date is realistic, we cut scope to fit it rather than sacrifice quality. If it isn't — I'll tell you upfront and propose a realistic alternative.
I offer ongoing post-production support: monitoring, bug fixes, feature development, dependency updates, and infrastructure maintenance. Most clients stay on a retainer after launch.
Telegram for daily communication, weekly calls for sync, GitHub/Linear for task tracking.
Absolutely. I frequently embed into client teams for architecture guidance, code reviews, and mentoring. I can also help you hire and onboard new developers.
It depends on request volume, data sensitivity, and how important independence from external providers is. For small volumes, cloud is faster and cheaper to launch. At scale, local inference pays back within months. I calculate the economics of both options against your actual numbers and help you decide without a predetermined side.
I work both hourly and on fixed scope. Hourly rates are in the senior-fullstack range and adjusted to engagement type (one-off audit vs ongoing development vs CTO retainer). For fixed-scope work, I always quote upfront after a scoping call. Billing is monthly for ongoing work, milestone-based for fixed projects.
