PORTFOLIO · MELBOURNE, AUS
AI Engineer
THE ANALYST BEHIND THE NUMBERS
I'm an AI and data specialist based in Melbourne. My work bridges statistical rigor and real-world application: from automation systems processing thousands of records to computer-vision models watching live traffic.
Data is more than numbers: it's how organisations understand decisions, improve systems and move with confidence. Every build starts with the same question: what is this data really telling us?
Beyond the datasets, you'll find me experimenting with new technologies and sharing what I learn: mentoring students, running workshops, and helping teams use data with confidence.
WHAT I BUILD WITH
Systems that reason over messy, real-world data and ship.
Models trained, tuned and measured until the numbers hold in production.
Teaching cameras to understand traffic, gestures and the physical world.
THE TOOLBOX
WHERE THE NUMBERS COME FROM
10+ automation workflows across Power Automate and Copilot Studio, lifting efficiency by 40%; invoice approvals cut from 2 days to under 2 hours; geolocation check-in app serving 450+ staff; onboarding admin workload down 30%.
15,000+ student records processed and 5,000+ assessments validated with 20+ staff, cutting errors by 25%; 10,000+ data points reconciled to deliver statewide exam results for 50,000+ students on time at 100% compliance.
Led a 12-member team building a YOLOv8 traffic system across 10,000+ images, lifting accuracy by 30%; managed 120+ hours of annotation and mentored 30+ peers to a 40% efficiency gain, delivering 100% on time and taking 1st place at Innofes 2024.
Predictive models over 30+ banking datasets cut loan misclassification by 35%; insights presented to 100+ consultants and directors nationwide.
Automated Power BI dashboards and SQL workflows across 6 branches: reporting time down 50%, tracking accuracy up 25%, KPI discrepancies down 30%.
Guided 30+ IT and data science students through programming workshops and career mentoring.
SELECTED WORK
Award-winning real-time traffic congestion detection using computer vision.
Detects vehicles, then tracks density and speed across 10,000+ images to flag congestion patterns for smarter urban routing. Built leading a 12-member team that earned 1st place at Innofes 2024.
Deep learning image classification achieving 98% accuracy using ResNet50.
ResNet50 + CNN classifying vegetables from still images and live video feeds, hitting 98% accuracy inside the first training epoch, built with smart agriculture in mind.
Real-time gesture recognition and virtual drawing powered by MediaPipe and OpenCV.
Hand-landmark detection turns any webcam into a drawing tool: sketch in the air, in real time. Open source with 18+ GitHub stars.
MORE BUILDS
Fiscal-year performance, profitability heatmaps and customer drilldowns for strategic decisions.
POWER BI · SQL · DAXWearable-driven supplement recommendations, People's Choice Award, VicHack 2024.
PYTHON · FLASK · POWER BIInterpretable stroke-risk classification with engineered features and balanced data.
PYTHON · SCIKIT-LEARNRandom Forest beat SVM, decision trees and neural nets on RMSE for urban energy planning.
PYTHON · REGRESSION MODELSPROOF OF WORK
AI system detecting modern-slavery risk in corporate transparency reports.
1st place nationally for a 95.93%-accuracy classification model on financial data.
Dual awards for the YOLOv8 traffic detection project and its presentation.
NutriSync, an AI-powered personalised health recommendation app.
Bachelor of Computer Science (Data Science)
High Distinction, GPA 7.0/7.0 · International Excellence Scholarship (25% tuition) · SEBE Peer Mentor · Business Analytics, Software Engineering & Vietnamese Student Clubs
Diploma of Information Technology
High Distinction, GPA 7.0/7.0 · Faculty assistant for the Data Science Concepts unit (SIT112) · Student Committee, organised 80+ attendee events
Google Advanced Data Analytics
Predictive modelling and data storytelling
Neural Networks & Deep Learning
CNNs and computer vision for real-world AI
SAS Programming 1: Essentials
SAS Base programming and statistical analysis
Ready to turn data challenges into opportunities? Whether it's analysis, visualisation or predictive modelling, let's talk.