Fullstack Developer & AI Integrator · TypeScript · Python · React · RAG · AI Agents
Fullstack developer and AI integrator. Build production-ready web apps, real-time systems, and AI-driven solutions — from concept to AI outcomes.
- 🤖 AI: RAG (pgvector), LLM agents (OpenAI/Claude), n8n automation, LangChain
- ⚡ Backend: NestJS, Fastify, Prisma, PostgreSQL, Redis, Python
- 🖥 Frontend: React, NextJS, TypeScript, FSD, TanStack Query
- 📱 Mobile: React Native (Expo)
- 🐳 Infra: Docker, Caddy, Cloudflare, GitHub Actions
🔭 Currently focused on docbrain — AI assistant for corporate docs (RAG + agents) and scan-agent — AI agent for freelance platforms.
AI / Automation
Backend
Frontend
Infra
| Project | Stack | Description |
|---|---|---|
| docbrain | FastAPI · Next.js · pgvector · n8n | AI consultant for corporate docs — RAG, agents, hybrid search |
| scan-agent | TypeScript · Prisma · PWA · LLM | AI agent for freelance platforms — parsing, evaluation, auto-reply |
| ai-automation-starter | Python · LLM · pytest | Modular AI pipeline template — data → AI → markdown → Telegram |
| corporate-transport | Fastify · React · Expo · Redis | GPS fleet tracking — monorepo with PWA + mobile app |
| support-ticketing-system | NestJS · Prisma · React | Helpdesk CRM with roles, comments, attachments |
| AnyWhereDesk | Docker · Caddy · Guacamole | Self-hosted remote desktop — SSL, DDNS, auto-backup |
I design and apply the ED Microservices Monorepo Pattern — an approach to organizing microservices in a single monorepo with clear responsibility boundaries.
📄 Full pattern description in Wiki
Core principles:
- Monorepo is not a monolith.
- Each microservice is an independent runtime/deployment unit.
- Internal service structure follows
app → features → services → shared. - Shared code lives in
packagesonly after proven reuse.