Machine Learning Engineer Sydney, Australia
I build AI systems that make it into the real world.
Production machine learning, computer vision, and agentic systems, grounded in careful research and measured by what they make possible.
A little context
Research-minded.
Production-focused.
I am a machine learning engineer focused on turning research into dependable AI products. At IHX, I re-engineered a document-intelligence product supporting around 90% of company revenue, reaching 99% table-classification precision and cutting pipeline time by half.
At the University of Technology Sydney, I research mechanistic interpretability for EEG foundation models. My work spans model evaluation, sparse feature analysis, computer vision, multimodal systems, and production AI engineering.
I build end-to-end systems, from data and model experiments through APIs, observability, and deployment. I also write about ideas and implementations I am exploring in the blog.
Selected work / 2025-26
Projects.
PitchSense
Computer vision for football tactics, turning ordinary match footage into pitch-level telemetry without GPS trackers or specialist hardware.
Explore projectPixel Sprites
Multimodal Memory & Agentic Content System
A multimodal AI application with specialized agents for understanding and creating content from images, video, and audio.
Explore projectHabit Tracker
AI-Powered Task Categorizer
A full-stack habit and task-tracking system that categorizes Google Tasks with AI and brings daily activity into view.
Explore projectAI Car Insurance Advisor
An AI insurance recommendation system combining driver-risk prediction, premium estimation, policy ranking, and evidence-backed explanations.
Explore projectEEG Foundation Model Interpretability
Research into how pretrained EEG transformers represent neurological states, using sparse feature analysis and activation-level interpretation.
Explore projectTools and methods
What I work with.
From model research and evaluation to the APIs, workflows, and monitoring around a production system.