UQ Social Timetable

Mobile App

• Developed a mobile application for students to view their own timetable alongside their friends' timetables.
• Published on the App Store and Google Play Store, reaching over 8000 users with 500 daily active users.
• Designed a user-friendly interface to display timetable information and allow students to manage their schedule. Initially written in Swift and later translated to Flutter for cross-platform compatibility.
• Leveraged Firebase's NoSQL cloud database to implement social networking features, enabling users to send, accept, and decline friend requests.
• Enabled real-time messaging and timetable sharing by utilizing Firebase's cloud database.
• Integrated OneSignal for push notifications for group chat messages and friend requests.

Maze Generation and Pathfinding Visualiser

Web App

• Developed a web application to visualise maze generation and pathfinding algorithms.
• Implemented maze generation algorithms such as Recursive Backtracking, Recursive Division, Prim's Algorithm, and Kruskal's Algorithm.
• Implemented pathfinding algorithms such as Greedy-Best-First Search, A* Search, Breadth First Search, and Depth First Search.
• Designed a user-friendly interface to allow users to select the size of the maze, the algorithm for maze generation, and the algorithm for pathfinding.
• Utilized React and canvas graphics with p5.js to create the application.

Neural Network Visualiser

Desktop App

• Built from scratch in C++ - complete neural network with custom matrix operations, backpropagation, and gradient descent (no ML libraries)
• Solves the legendary XOR problem that stumped early AI researchers, demonstrating non-linear pattern learning through hidden layers
• Real-time SDL2 visualisation shows decision boundaries evolving as the network trains, making AI learning visually intuitive
• Features three classic machine learning problems: XOR logic gates, circle boundary classification, and spiral separation with configurable network architectures
• Live training metrics display current epoch count, error rates, and learning improvement to track neural network convergence in real-time

Sorting Algorithm Visualiser

Desktop App

• Developed a desktop application to visualise sorting algorithms.
• Implemented sorting algorithms such as Bubble Sort, Selection Sort, Insertion Sort, Merge Sort, Quick Sort, and Heap Sort.
• Designed a user-friendly interface allowing users to select the size of the array, the sorting algorithm, and the speed of the visualisation.
• Utilized Java Swing to build the application.

Snake Game

Desktop App

• A simple window-based Snake game implemented in C using the SDL library.
• The game features a snake that moves around the screen, eating food to grow.
• The game ends when the snake collides with the walls or itself.
• Written in C using the SDL library for graphics.