Mahadev
Kumar
I build fast, beautiful web apps and intelligent systems — full stack, from the pixel to the API to the model.
About Me
Crafting web experiences
that matter
I'm a Full Stack Developer with a focus on the MERN Stack and AI integration. Expert in building scalable web applications using React.js, Node.js, and MongoDB, with a specialized background in developing RAG systems and LLM-powered automation. Proficient in CI/CD via GitHub-Actions and Computer Vision (OpenCV). Strong foundation in Data Structures and Algorithms with a commitment to writing clean, optimized, and maintainable code.
Alongside core web work, I bring hands-on Machine Learning knowledge to production — recommendation engines, NLP pipelines, and data-driven features.
Fluent in C++, Python, and JavaScript. My competitive programming roots give me a sharp algorithmic edge on hard problems.
Performance First
Core Web Vitals, optimal bundles, and smooth 60fps UX on every device.
ML-Integrated Web Apps
Bridging web dev with machine learning — recommendations, NLP, and smart features in production.
Security & Quality
Clean, tested, documented code that teams are proud to maintain.
Ship Fast, Ship Right
Concept to deployment — full lifecycle ownership, on time without cutting corners.
Technical Expertise
Skills & Technologies
Frontend
Backend
Languages
ML / AI
Selected Work
Projects & Work
Campus
Management System
A comprehensive management platform designed to streamline campus operations. This project showcases a scalable MERN architecture, focusing on optimized frontend delivery and secure data handling.
Laundary Mart
Laundry Service Management
Developed responsive web application for laundry service management using HTML5, CSS3, and JavaScript. Implemented mobile-first responsive design, interactive user interface, form validation, and image gallery integration. Organized modular code structure with semantic HTML5 markup for accessibility and SEO optimization.
Vehicle Predictor
Vehicle Classification System
Engineered deep learning vehicle classification system using PyTorch CNN architecture. Built FastAPI server integrating pre-trained neural network model for real-time vehicle predictions. Developed responsive web interface with HTML/CSS for image upload and classification. Implemented complete ML pipeline from image preprocessing to model inference with production-ready deployment.
Learn-To-PR
Open Source Project
Learn-To-PR is an open-source web platform designed to help beginners start contributing to open source projects easily. The platform provides a structured environment with categorized issues, automated pull request labeling, and clear contribution guidelines to reduce the entry barrier for new developers.
Portfolio
My Portfolio
Portfolio
Get In Touch
Let's build something remarkable
Send a message
Your message is stored securely in our database.
Direct Email
Availability
Response Time
Typically within 24 hours on weekdays.
Find Me Online