I’ve built the thing I’m now asked to judge.

Too early

In 2011 I was the founding CTO of WhipCar, the UK’s first peer-to-peer car-sharing marketplace. Strangers renting each other’s cars, by the hour, with the whole thing coordinated online.

The idea was right. The market wasn’t ready — smartphones weren’t universal, the insurance products didn’t exist, and the behaviour change we needed was a decade away. It didn’t work.

It’s the most useful thing on this page. Being early is indistinguishable from being wrong, right up until it isn’t, and I now recognise the shape of that argument when a founder makes it to me.

Neutron Creations

I co-founded a web and digital consultancy, which is where I learned to run technical work as a business rather than as a craft — scoping, pricing, and delivering to someone else’s deadline.

The right time

When the market had matured, I was recruited as founding CTO of the peer-to-peer and managed car-sharing platform Hiyacar.

I built the platform, the engineering team, and integrated the keyless access and telematics system that made unattended vehicle handover work. Without it, every booking needs the owner and the renter in the same place at the same time. The platform reached 250,000+ members.

I later became interim CEO. Running the whole business is the reason I can tell you what the rest of a company needs from engineering and why it usually doesn’t get it.

AI that had to pay for itself

I put machine learning and LLM-driven systems into production where they had to justify their cost — fraud detection, and operations work that had previously needed people — rather than as a demo for a board deck.

I’m not a research engineer, though. I’ve led the ML work and shipped the systems; if you need a novel model trained, that’s a different hire.

Now

I’m building Julie AI, and I’m engaged with a mobility startup.

Alongside that I take a small number of fractional CTO engagements and technical due diligence instructions — a maximum of two at a time, which is the honest limit for work that depends on my attention rather than my availability.

I don’t take due diligence engagements where there is a conflict with my current engagements. That’s set out in full in the conflicts policy.