An independent AI practice. Built on curiosity.
Most AI projects begin with a big plan. I'd rather begin by building. We try something small, find out where it really hurts, and build the thing that genuinely helps. Often that's an agentic system that takes real work off people's hands, and it should feel effortless, even quietly delightful, to use.
How I work
No big decks. No six-month roadmaps. Just a fast, honest loop.
I dig into what's actually going wrong, with you. It's usually hiding somewhere other than where everyone's been looking.
A small working version teaches us more than any plan. We find the truth by trying it for real, then changing our minds quickly when we're wrong.
Then I build the proper thing. Intuitive, quiet, and easy. The kind of software people are quietly glad to open.
Things I've built
Some for clients, some just because I was curious. Have a look, and hover to see a little more.
For clients
Out of curiosity
Snap any ingredient list and get an instant gut-health read. It flags what matters and tells you why.
An IFS therapy voice companion. Talk to your inner parts and notice the patterns over time.
A habit coach that lives inside WhatsApp. Daily check-ins, a good memory, and no nagging.
Personalised bedtime stories in seconds. Pick a storyteller voice, add a few magical keywords, and get a tale that couldn't exist without this exact moment.
A playful virtual fish that chases lasers, begs for food, and brightens your day. Sometimes the best experiments are the silliest.
How hard were your commits, really? It reads your git history and scores the effort.
Strength training without the guesswork. Personalised programmes and sensible progression.
A bit about me
For about ten years I've worked in business transformation, a lot of it at Deloitte. That mostly meant getting in amongst how a business actually runs, working out where it was stuck, and helping people decide what genuinely needed to change before anyone built a thing. Later I moved into building too, including generative AI products in their innovation team.
What I slowly realised is that the code was rarely the hard bit. The hard bit was figuring out what was actually worth building, and then making it something people didn't mind using. Most of that I learned by sitting with teams and watching how they really work before touching anything. I still start there.
Now I build AI products, more and more of them agentic, for myself and for the people I work with. What I enjoy most is taking a team from "we should probably be doing something with AI" to a real working tool in a matter of months. I'm curious, I build fast, and I care a lot about the small things that make something nice to use.
Hom
Established business or scrappy startup, if you can feel that AI should be changing how you work but you're not sure where to start, that's my favourite kind of conversation.
Start a conversation →