Thought piece

Built for a world that moves

Neil Middleton · 16 October 2026

A product is a bet that the world holds still. I made that bet once. It held for eighteen years, because for eighteen years the world I had bet on held still enough.

The previous generation of this software, the one I founded and ran before 2018, was an all-in-one workflow system for small regional airlines, the third and fourth tier of the industry. It ran most of a small carrier's operating day on one monolith over one relational database, with an organisation that had grown to fit the same market. It fitted its market well, and that was the problem. The market was small fleets, tight margins and a supply of small aircraft that has been contracting for decades. Moving up to the large carriers would have meant a different scope, a different stack and a different organisation, which is to say a different company. Every part of the product was fitted to the tier it served, so no part of it could leave. Fit is a moat until the world moves. Then it is a wall.

Two things are moving now, and a product fitted to a still world meets both at once. The market is moving: the small-aircraft world keeps contracting, and the operators who remain consolidate. The interface is moving: booking is leaving the form over a database and arriving as an agent working an event stream through an API. A static product can follow neither. It can only be what it was fitted to be, in a world that has stopped asking for it.

UnityTrip was built on the other side of both moves, and in a different market. The organisations that fly their own workforce, in energy, resources, humanitarian work and field operations, are not a marginal market and they are not shrinking. Their travel does not depend on who is manufacturing small aircraft this decade. It depends on people needing to be at a site, a platform or a camp on a date, on whatever carries them: a leased turboprop, a chartered helicopter, a crew boat, a commercial flight. The mechanism that governs that movement, one record, two gates and a policy that rations a seat with no fare, is the same whichever asset carries the person. That market is large, it grows with the operations it serves, and it pays for governance because the alternative is spend that nobody can reconcile.

So the engineering is different in kind. One event-sourced record rather than a schema each customer bends. Configuration rather than code, so a new operation is a new policy document, not a new build. An event stream and an API that an agent can work, with frontier AI at build time and determinism at run time, so the booking path stays auditable while the interface around it changes.

And the culture is different in kind, which matters more than the stack. We sell an enquiry before we sell a treatment: the review reads an operation's records before anyone proposes anything. We have shipped forty production changes over a week when an operation needed them, and retired the ideas that did not swim. We publish the mechanism and the evidence so the next operation can check it rather than take our word. None of that is a policy we wrote down. It is what people do when they are interested in what they have not seen yet.

That is the bet. Not that I know where the market or the interface goes next. I do not. The bet is that a mechanism built to be configured, a stack built to be driven, and a team that wants to find out can move with them.

The world will move again. The question for any product is whether it can move too, and whether the people behind it still want it to.

Also: The capacity to surge, Many presentations, one mechanism, The second shift and UnityTrip and SAP: a single source of truth. The investor thesis sets out what this means for the company, and the platform thesis what it means for the product.