[ MEHDI.TERZI / FULL-STACK ]
creative web
developer.
Websites, web apps, e-commerce, SaaS & AI-powered products — end to end.
Products, systems and experiments connected across one living field.
// PROCESS
01
Interactive Sites
Branded sites with motion as a first-class material — cursors, transitions, scroll choreography.
02
Animated UIs
Productized interfaces where motion communicates state, not just decorates it.
03
Full-Stack Web Apps
End-to-end products: schema, API, UI, deploy — most often Nuxt + Supabase + Vercel.
04
Performance Hardening
Existing sites taken from "fine" to sub-1.5s LCP without losing the craft.
Local LLM
LOCAL AI
Qwen-family models running fully local — private inference on my own hardware.
Business Agents
AI AGENTS
Local automation agents that run the repetitive back-office of small businesses.
Virtual Try-On
COMPUTER VISION
Image-transformation research with ComfyUI — putting products onto people.
Generative Video
GENERATIVE AI
Wan / ComfyUI pipelines for short-form generative video.
Subtitle Engine
AUTOMATION
Automatic transcription plus styled, burned-in subtitle rendering.
Moto Telemetry
HARDWARE
CAN bus + Raspberry Pi experiments reading a motorcycle in real time.
Spatial Canvas
WEBGL
The DOM + WebGL hybrid engine this very space runs on.
Self-Hosted Edge
INFRASTRUCTURE
A homelab that serves production — Docker containers behind a Cloudflare tunnel.
// ABOUT
Mehdi Terzi
Istanbul, TR
I design and build the kind of websites people remember — sites with motion, weight, and a point of view. Brands, studios, hotels, retailers, law offices, products, internal tools.
I work end-to-end: design, frontend, backend, deploy, and SEO. From a single landing page to multi-tenant SaaS, AI-powered management systems, and self-hosted infrastructure — everything on this canvas is live and in production.
// CONTACT
[email protected]0539 373 7901- GITHUB@mehditerzi ↗
MEHDI.TERZI
SPACE 01
MEHDI // VOID
FORMING THE SPACE