I trained as a scientist, an MSc in statistical genetics at the University of Cape Town. I spent my time using statistics to work out how species evolved using information encoded in their DNA. That kickstarted my lifelong passion for using data to understand how complex systems work. Then I spent a few years consulting in the private equity and investment space and in the insurance industry.

After that, I co-founded a digital healthcare business called Guidepost and spent fourteen years on it. We built the software, data and compliance infrastructure that helped deliver skilled diabetes nursing to thirty thousand patients, with quality we could measure and stand behind.

These days I operate as a solo consultant and focus on two things that I enjoy: writing code to bring practical AI systems to life, and sitting with the people running the business to solve practical commercial problems.

There are three ways I serve my clients:

Retainers. An ongoing relationship with a leadership team. I connect more of their data to AI, build what is useful, and we work out together where it helps most.

Projects. A defined piece of work with a deadline and a fixed price. The kind of thing that takes the rough idea of "we should try AI for this" and gets it actually running and useful. Building software, figuring out how to get messy business data organised and structured etc.

Partnerships. I have teamed up with a handful of companies in various industries where I like the people and the market. We figure out together how to use AI to take the industry forward. They bring their market insight and data. I bring the science and the code. I currently focus on a few professional services verticals.

I write weekly on the Phase Transitions AI Substack and on LinkedIn, from the business view at the top and down to the deep technical stuff underneath.