with commercial experience in business systems, data analytics, reporting, and machine learning. Passionate about transforming data into actionable business insights using Python, SQL, Power BI, and modern analytics techniques.
I'm a BSc Computer Science graduate from IU International University of Applied Sciences, specialising in machine learning and data science. My thesis benchmarked XGBoost against deep learning approaches for network intrusion detection. With a background in business operations and IT support, I understand how data fits into real business contexts. Based in Gauteng, South Africa, and open to data analyst, BI developer, and data science roles locally and fully remote.
My thesis focused on evaluating the performance of different machine learning algorithms in detecting network intrusions. I conducted a systematic literature review to identify the most effective aproaches. In addition I added my own XGBoost implementation to benchmark against the literature. The results showed that LSTM algorithms performed the best using the UNSW-NB15 dataset. My binary XGBoost implementation achieved the best results on the CIC-IDS-2017 compared to other XGBoost implementations in the literature. The multiclass XGBoost implementation achieved good results compared to the same algorithms on the CIC-IDS-2017 dataset. This research contributes to the field of network security by providing insights into the effectiveness of machine learning techniques for intrusion detection and highlights the potential of XGBoost as a viable option for real-world applications.
The thesis is available to read by clicking the "Read Thesis" button. The code is also available on GitHub. More details can be found in the projects section.
Classify network traffic as normal or malicious using machine learning to detect cyber attacks using the CIC-IDS-2017 dataset.
Designed and built a fully normalised relational database for Stayalot, a fictional short-term accommodation platform. The schema spans 20 interconnected tables with foreign key constraints, enum types, and privacy-separated data, supporting multi-table analytical queries across bookings, payments, and host/client relationships.
Built a full-stack habit tracking application with a Python frontend and PostgreSQL backend, featuring persistent data storage, user metrics tracking, and summary reporting through a custom-designed schema.
Developed a full-stack Hangman web application using Node.js, HTML, CSS, JavaScript, and with MongoDB for persistent game state storage and a RESTful API architecture, hosted on a local server.