Projects and Activites
AWS DeepRacer First Place
We had a fantastic opportunity to delve into the topic of AI at Amazon's DeepRacer event.
The goal of the competition was to train our AI models to navigate as quickly as possible around a track with total reliance on the AI model.
We had three weeks to plan, develop, and test a learning algorithm for our AI model.
Weeks of training, model adjustments, and virtually testing the lap time later, my partner and I were invited to the finals held at the University of Sydney.
With a laptime of 7 seconds, we finished first out of 40 participants
and were awarded a very generous $250 worth of Amazon merchandise.
This event was also held at a nearby university 'University of Technology' where we beat their winning time of 13 seconds by almost half!
Distributed Machine Learning Model (Federated Learning)
This project showcases the use of Machine Learning, specifically Federated Learning (FL) to predict housing prices using the California Housing Dataset.
Within this project you will see that training is done in a distributed manner where the dataset is split and distributed to 5 different clients.
The clients will then train their designated piece of data and sent it back to the server to be aggregated to this main model,
and this process will be completed N number of times.
This manner of distributing to clients is extremely effective in terms of communication costs as all training of the data is done locally and only the model parameters will be sent back to the server,
meaning enhanced security and privacy as no datasets are ever being transmitted.
*Please check Readme.txt for program instructions*
Automated Web Vulnerability Scanner – Python CLI Tool (Capstone Project)
Developed a Python-based command-line tool that integrates the OWASP ZAP API to automate web application vulnerability scanning. The tool detects critical security flaws such as SQL injection, cross-site scripting (XSS), and authentication bypasses. Key features included adjustable scan strength levels, real-time progress feedback, and color-coded output to highlight the severity of detected issues.
Collaborated in an Agile development team, conducting regular sprint meetings, client review sessions, and live demos to validate functionality and incorporate stakeholder feedback. Final deliverables included a professionally styled HTML report summarising findings, with an AI-generated executive summary outlining key threats and tailored remediation recommendations for non-technical stakeholders.
The project emphasised secure coding practices, usability, and automation—helping streamline web app security assessments in a scalable and client-friendly format.
Due to confidentiality reasons I am unfortunately unable to link or provide any source code.
Airport System [Adjacency List Representation]
In order to deepen my understanding of data structures, I constructed an airport system using an adjacency list representation as I was having trouble completely comprehending graphs. The program gives the user complete control over features including showing the number of airports, adding airports, adding flight paths between airports, getting all flights departing from a particular airport, and listing all airport names.
Stock Matching Engine
Trading and the stock market are great interests of mine. I wanted to incorporate one of my favourite abstract data structures into this project because I'm intrigued by development from a trading perspective. The engine employs the 'Priority Queue' ADT, and the management of buy and sell orders is done using min and max heaps. A trade is carried out and executed if the asking buy order value is more than or equal to the selling order value, which is determined by comparing the min & max of the heaps.
Game Cheat Trainer
This project was a cheat trainer for a game called "Faster Than Light" that I made in C#. The program consisted of lots of memory manipulation to write my own memory to pointers which would change the values of variables within the games code to take control of the game.
Along side coding the program, I created a GUI using forms in Visual Studio to make the program more user friendly.
This Website!
Coding this website has increased my confidence in HTML and CSS tremendously! I also really enjoyed delving into the design aspect of programming.
A lot of the knowledge used to create this site was gained at a workshop hosted at the Google office in Sydney. It was an amazing experience, and I look forward to many more through my university's very own 'Google Student Development Club'.
