Now

What I'm doing now.

Last update: 22 September 2026, from Nutrify HQ, Brisbane, Australia

Started a new job

In April, I started a new job as Member of Technical Staff at Artificial Analysis.

My friend Michael was working there and we were talking and he said I'd probably be a good fit.

It's a crazy fun time!

We partner with and evaluate all of the largest AI labs in the world.

I'm in charge of the Search Index (evaluating how well a search API works to retrieve information for an agent).

Half our team in based in Australia (me included) and half is based in San Francisco.

Launched Nutrify 3.0

My brother and I just launched a biggggg update to Nutrify, our app dedicated to helping people focus on eating more whole foods.

Inside you'll three new ways to track foods. Now you can snap it (photo), text it, say it or scan it, add stickers, text to correct, set favourites and much more!

Read more in the Nutrify 3.0 launch blog post or download the app on the App Store.

Happy whole fooding!

Screenshots from Nutrify 3.0. **Left:** Take a photo of your meal and Nutrify will break down into ingredients with weight and serving size estimates. **Middle:** No photo? No worries, just text or say what you ate. **Right:** Add stickers to your meal with nutrition summary overlays and share it with friends. There's over 1000 custom designed colourful food stickers.

Getting fit

My wife and I would like to start a family soon.

So I'm doing my part by keeping my health in check.

Mostly by removing things.

No alcohol, no junk or processed foods (eating mostly whole foods with Nutrify of course).

Ticking off all the basics: moving every day, plenty of walking, weight lifting, whole foods, enough sleep.

I started the year at 87kg with a bit of Christmas feasting fluff, as of September 1st, I'm about 81-82kg. I'd like to hover around 80kg, I think that's a perfect weight for me.

Launched a new course on Hugging Face

In my AI and machine learning workflows, I use Hugging Face basically every day.

I had a consulting contract recently and I used Hugging Face for every step of the build process from model to demos.

This new course is designed to teach people how to create their own models.

We'll follow the process of:

Data – Getting a dataset for your problem.
Model – Choosing an existing model on Hugging Face to customize to your own problem set.
Demo – Creating a sharable demo of what you've done so other people can try it out.

All of the code and text for the course is available for free at learnhuggingface.com and you can sign up to step-by-step video walkthroughs on Zero to Mastery (all videos are 100% taught by myself).