Featured Projects
Production platforms I've helped build and ship
Huubr — Business Directory Platform
Contributed to a MERN-based business directory platform serving 800+ business profiles. Developed RESTful APIs and integrated dynamic search, filtering, and profile management features. Built reusable React/Next.js components to deliver responsive, SEO-optimized user experiences. Improved Core Web Vitals (LCP, CLS) and collaborated with cross-functional teams to deliver production-ready features.
Idris Book Bank — Online Book Store
Built and maintained the frontend and backend for a production e-commerce platform serving books, stationery, school bundles, and product bundles. Implemented category browsing, product search, cart flows, and inventory-aware product listings for a Rawalpindi-based retail business.
Athlena — Modest Activewear Store
Developed the frontend and backend for Athlena, a production e-commerce storefront for modest activewear and loungewear. Built shopping flows with cart and wishlist, customer account features, product browsing, and a responsive retail experience under the Attire Central brand.
VATT — Virtual Academy for Teacher Training
Led DevOps, CI/CD, and platform architecture for VATT Balochistan — a government teacher training portal serving 10,000+ registered teachers. Designed containerized deployments, configured Redis for session management, handled database restores and querying, and resolved production PHP backend issues to keep the LMS stable at scale.
Omarchy TTS — Offline Text-to-Speech
Built a system-wide offline TTS tool for Omarchy on Hyprland using Piper. Highlight or copy text, press Super + Ctrl + R, and hear it read aloud with a GTK overlay that highlights words in real time. Includes 16 curated offline voices, speed controls, save-and-resume sessions, and a warm Piper daemon that cuts first-speech latency to under 350 ms.
Urdu Transcriber — YouTube Urdu Transcription
Built a Streamlit app that downloads YouTube videos and transcribes Urdu audio using OpenAI Whisper, used to process 700+ Urdu videos at scale. Supports multiple model sizes, batch processing across large workloads, and rich export options including timestamped Word documents with clickable links back to the source video.