Work

Selected projects built with production-minded discipline.

AutoTM

Turkmenistan's vehicle marketplace rewrite — an active Turborepo rebuild with Expo, NestJS bounded contexts, Prisma, and air-gapped deployment constraints.

Expo SDK 55
React Native 0.83
NativeWind v4
NestJS 11
Fastify
Prisma 7
PostgreSQL 16
Redis 7
BullMQ
Next.js 16
Tailwind CSS v4
shadcn/ui
MinIO
Sharp
Zod
Socket.IO
Turborepo
Docker
Caddy

Active ground-up rewrite of Turkmenistan's vehicle marketplace. S1–S3 are shipped; S4 Listings CRUD is the current build slice. The architecture is designed for air-gapped deployment with no cloud dependency on the critical production path.

Personal Engineering Portfolio

The portfolio you are viewing — a live meta case study in delivery discipline, Railway deployment, GitHub Actions gates, Docker, Prisma migrations, and private client rooms.

Next.js 16
TypeScript
Tailwind CSS v4
shadcn/ui
Prisma 7
PostgreSQL
Auth.js
Vitest
Playwright
Docker
GitHub Actions
Railway

138 tests, GitHub Actions quality gates, WCAG 2.1 AA accessibility checks, Dockerized Railway deployment, startup Prisma migrations, custom domain configuration, and private client rooms. The portfolio proves the engineering system it describes.

MyORL — ENT Clinic Platform

Bilingual healthcare website for a private ENT surgical clinic in Athens — Next.js 16, Strapi 5 CMS, Meilisearch, fully deployed and serving patients.

Next.js 16
React 19
TypeScript
Tailwind CSS v4
Strapi 5
PostgreSQL 18
Meilisearch
Docker
Caddy
Vitest
Playwright
Railway

Live bilingual healthcare platform for an Athens ENT clinic — Next.js 16 frontend, Strapi 5 CMS with full content handover, Meilisearch patient-facing search, and Playwright-tested booking flow. Deployed on Railway and serving real patients.

Gonka — AI Inference Infrastructure

Specified, deployed, and supported a 24-node GPU cluster (192× RTX 4080) serving open-weight LLMs on the Gonka decentralized AI inference network — plus an OpenAI-compatible API gateway built on top of it.

GPU Infrastructure
RTX 4080 Clusters
Open-Weight LLMs
Qwen
vLLM
Decentralized Inference
Express
TypeScript
Zod
SSE
Docker
Vitest

24-node / 192-GPU inference cluster specified, deployed, and supported for 3 months on the Gonka decentralized AI network, serving Qwen 235B-class models — plus GonkaProvider, a public OpenAI-compatible API gateway with streaming validation and an upstream vLLM fix.