Deep-Dive Showcase

Detailed Review of Every Project

Explore thorough breakdowns for each work in this portfolio—including complete descriptions, live production deployments, source code repositories, image galleries, and full technology stacks.

Total Projects

Documented in portfolio

Featured Works

Highlighted flagships

Technologies

Frameworks, DBs & APIs

Live Demos

Active interactive URLs

Jump to:

Filter:

#01★ FeaturedLive Online

Dedicated Page →

1 / 3

AkarSham (عقار شام) — Real Estate Marketplace Platform

Project Overview & Review

A full-stack, multi-app real estate marketplace built for the Syrian market, supporting property listings, agent verification, push notifications, and a fully admin-configurable content/field system — across a native mobile app, a public marketing site, and an internal admin dashboard, all backed by a shared REST API. Architecture Overview AkarNow lets users browse, search, and filter property listings (sale/rent) on an interactive map, save favorites, and list their own properties. Real estate agents/offices go through an ID-verification flow to get a "verified" badge. Admins manage users, listings, cities, verification requests, dynamic property fields, and even the public landing page content — all from a dedicated dashboard, with real-time stats and charts. Four independent applications sharing one backend API: Mobile app — customer-facing Android/iOS app (Expo/React Native) Admin dashboard — internal operations panel (React/Vite) Landing page — public marketing site (React/Vite) Backend API — shared REST API (Node.js/Express)

Technical Details & Status

Deployment Status: Production Live
Source Repository: Public on GitHub
Screenshots & Assets: 3 Images
Stack Count: 7 Core Technologies

Launch Live Demo View Source Code Single Review View →

#02★ Featured

Dedicated Page →

RagChat App

Project Overview & Review

A local-first RAG document assistant built with Next.js, LangChain, Ollama, Clerk, Neon PostgreSQL, and pgvector. Users upload TXT, PDF, and image files, then receive streamed, context-grounded answers based only on their own documents.