Stones Corner AI
Social robots for aged care, which I co-founded. I built the software side: a real-time voice agent, the robot's on-device services, safety monitoring for every conversation, and the cloud and apps around it.
Projects
Foraging Lab
A live tank of evolving spiking-neural-network creatures you can poke at, with an inference API and a model registry behind it. Organisms that eat survive and breed, and the best brains are benchmarked and promoted automatically.
CompCase AI
Sentencing research for Queensland criminal lawyers, which I co-founded. An LLM pipeline turns court judgments into cited, structured data, and a practising lawyer's golden set checks every prompt change.
spike-forest
A small, modular spiking-neural-network simulator built around one question: what happens when several plasticity rules run at the same time in the same network? It's the engine behind Foraging Lab's brains.
Spiking neural networks in TensorFlow
An early side project: a spiking network simulator with plastic synaptic delays and weights.
About
In my PhD at the University of Queensland I worked on unsupervised learning in spiking neural networks with varying types and intensities of noise. Specifically, I studied how the plasticity of action potential conduction delays can be used as a functional computational mechanism, similar to weight learning in networks of spiking neurons.
I've also been principal data scientist at Vincents, a national professional services firm, where I led the team that built LLM tools for forensic accountants and lawyers. Before that I was a postdoc at the Queensland Brain Institute, where I built a pipeline that processed hundreds of terabytes of brain imaging.
I've taught at the University of Queensland for over a decade, from tutoring to running 500-student courses, and I supervise student projects there. In my spare time I like to try out new hobbies like watchmaking and 3D printing, or take my dogs out adventuring.