Final Year Capstone project in collaboration with Transport NSW & CLOCS-A. Initiative to develop several-cutting edge apps that help construction and logistics companies across Australia monitor and adhere to compliance standards in trucking logistics. Collaborated as a team of 4 developers to build 3 fully integrated apps.

Procedurally generate game assets using artificial intelligence, custom pipelines and natural language processing (NLP). Create game assets from simple prompts such as a game description and theme. Utilises dreambooth research from Google and custom trained deep learning text-to-image generation models.

Built a custom neural network model that can be trained via API. The idea behind it was initially a chatbot which evolved to become a NPC creator for games. The API was built using FastAPI and SQLAlchemy. The neural network was built using PyTorch, utilising 3 linear layers and a relu activation function.

Custom game project showcasing the implementation of different software design patterns and data structures built in C++. This includes the state, composite, component, command and factory design patterns, as well as a graph/node data structure to represent the game world with different locations with adjacent location nodes.