Team project part of a submission to Level Up Showcase 2025, Moodcaster is a 2D/3D alt-controller dungeon-crawling horror game where the player casts spells with their emotions using the webcam. The implementation of Morphcast, the artificial intelligence used to enable efficient and accurate facial recognition, was the core focus of the 4-month project.
Efficient data transfer through custom web socket server
Designing of particle systems and shaders
Handled critical connections between art and code
Created AI controllers for various enemy types, using a custom state machine system
Solo passion project designed to be a modular, reusable solution to generating room-based maps in C#. Supporting up to 40 rooms per floor, on a 13x13 grid, generation uses a nearest-neighbour approach, and allows for Binding of Isaac inspired secret room selection, special rooms, and keeps loops to a minimum.
Gained proficiency in algorithm optimization, allowing for faster and more memory efficient generation
Modular and generic, designed to be inherently reusable
Assisted in learning about node-based AI pathfinding algorithms, such as Dijkstra's, A*, and nearest neighbour
First year team project, this procedurally generated dungeon-crawler project was created to build teamwork and programming skills, completed over the 12-week semester.
As lead programmer, I was tasked with ensuring code quality was brought to standard, and all systems created by both myself and my colleagues we're completely functional
Assisted in design choices, such as overall theme, sprite assets, and shaders
Helped colleagues learn about general coding standards and practices utilizing background knowledge of Unity and C# development