Future

For decades, making a program mostly meant writing it: line after line, by hand, for months. The cost of that work decided which products could exist and which could not.

Today AI models write code, test it and fix it at a pace no group of people can keep. Writing has stopped being the bottleneck.

A machine does not know what a beach club needs in mid-August, or what the person planning warehouse shifts needs. Understanding the problem, choosing what to build and what to leave out, noticing that something works but is wrong: that remains human work.

Responsibility stays human too. Behind every Kodnia product there is a person who has read, tested and approved it, and who answers for it.

When building costs less, it makes sense to write software for precise needs: one trade, one kind of company, one problem. Products that used to be not worth starting, because the market was too small to pay for a whole department.

This is where we want to be: many specialised products, each done well for the people who use it, rather than one huge program that does everything halfway.

We want to stay a small, very capable team, and grow the number of products, not the number of desks. Each new product uses what we learned building the ones before.

We want automation to go all the way: not just writing the code, but testing it, releasing it and keeping it tidy, with a person checking every step that matters.

And we want it to stay clear who decides. Machines are a formidable tool; the direction comes from people who know the work.