Sign Language Recognition System
A real-time hand gesture recognition system for Devanagari script using MediaPipe and SVM, built to translate sign language into readable text.
I'm Pratik Devkota — a MERN stack developer and AI enthusiast focused on building practical applications. From real-time chat systems to sign language recognition, I like working on things that solve real problems and actually run in production.
Selected work
A streaming analytics platform processing for B2B ecommerce platform.
A real-time hand gesture recognition system for Devanagari script using MediaPipe and SVM, built to translate sign language into readable text.
A full-stack chat platform with live messaging, user authentication, and socket-based communication using the MERN stack.
A food delivery system with multi-role support including admin and restaurant owners, featuring dynamic menus and order management.
A scalable online store with product management, authentication, and structured backend APIs.
improved user experience and accessibility for the Hajurbuwa platform with a React Native mobile application, featuring real-time order tracking and push notifications.
A modern developer portfolio built with Next.js, focusing on clean UI, performance, and responsive design.
About
I’m Pratik Devkota, a Computer Engineering student and full-stack developer based in Nepal. I started with web development and gradually moved into building complete systems — from backend APIs to real-time applications and machine learning projects.
I enjoy working on practical problems. Whether it’s building a chat system, designing an e-commerce backend, or experimenting with computer vision, I focus on making things that actually work, not just look good on paper.
Currently, I’m exploring AI and improving my backend skills while building projects that combine real-world use cases with solid engineering practices.
Stack