Approach
A business first. The technology serves it.
Five steps, in order — and the order is most of the value, because it settles the commercial questions before a line of code makes them expensive to revisit.
Start with the gap, not the idea
An existing market where the need is real and current technology serves it badly, or not at all. That distinction matters: inventing demand is the hardest thing in business and we do not attempt it. We look for people already trying to do something, working around tools that were never built for them.
Work out how it makes money
Before anything gets built. Who pays, what for, how often, and what it costs us to serve them — because a gap in a market is not automatically a business, and most of them are not. Getting this wrong is not a setback you engineer your way out of later; it is the difference between a company and an expensive hobby.
Use technology to keep it simple
Technology is the means here, not the point. It is what lets a very small company serve people properly: no sales calls, no onboarding, no back office. Most of these gaps stay open because the conventional answer would be too heavy to be worth anyone's trouble — so if the thing we build needs explaining, we have not finished.
Stay light enough to be wrong
A venture that needs a large team or a large bet before anyone can tell whether it works is the wrong shape. We keep them cheap to start and lean to run, so an idea can be tested honestly and closed without drama. Being wrong quickly is affordable. Being wrong slowly is not.
Then run the business
Launching is not finishing. Running it means the support, the disputes, the refunds, the compliance and the tax — and it is what makes every earlier decision honest, because nobody cuts a corner on refund handling when they will be the one processing the refund. It is also where a business either works or does not.
What we don't do
We do not take client work and we do not build technology for other companies. Every venture here is one we own and are still on the hook for.
We do not chase markets that need to be created, and we do not enter one just because it is large. If the need is not already there, we are not the ones to invent it — and if we cannot see how it makes money, it does not get built.
And we do not run more than one build at a time. That is a real constraint and it costs us speed, but the alternative is two half-finished ventures and no operating business.