// Computer Systems Engineer
Final year Computer Systems Engineering student at the University of Sunderland, UK. Building full-stack applications, Android apps, and AI-driven systems.
// About Me
I'm Thivekshan Rajakumar, a passionate Computer Systems Engineering student originally from Sri Lanka, currently in my final year at the University of Sunderland, UK — on track for a First Class degree.
My journey started with learning C, Java and Python, building my first Android game from scratch, then progressing to full-stack web development with the MERN stack. I have since completed a Data Analytics certificate, earned my AWS Cloud Practitioner certification, and am now building real-world AI and computer vision systems using PyTorch and TensorFlow for my final year project.
I am a curious, self-driven builder who loves solving real problems with technology. I have 3 years of freelance web development experience and have built and deployed multiple projects independently. I am actively seeking graduate roles and internships in software engineering, AI and data — ready to bring real value from day one.
// Skills
// Projects
A fully functional Flappy Bird clone built natively for Android using Java and Android Studio. Designed and implemented game physics, collision detection, score tracking, and a complete UI from scratch — demonstrating core mobile development and object-oriented programming skills.
A fully responsive personal portfolio website built from scratch using pure HTML, CSS and JavaScript — no frameworks. Features smooth scroll animations, a custom cursor, interactive sections and is deployed live using GitHub Pages. Currently the site you are viewing!
A full-stack web application built using the MERN stack. Features include user authentication with JWT tokens, a RESTful API backend built with Express and Node.js, a dynamic React frontend with component-based architecture, and a MongoDB database for persistent data storage.
A deep learning image classifier built with TensorFlow and MobileNetV2. Classifies images into 1000 categories with confidence scores using a pre-trained CNN model trained on 1.2 million images.
Final year capstone project — a computer vision system using Convolutional Neural Networks with PyTorch for image classification (MNIST) and real-world object detection on the KITTI autonomous driving dataset.