Most company pages tell you who they are for. Nobody believes that copy, because there is no decision behind it: if it suits everyone, it tells you nothing.
So we wrote the opposite instead. Five situations where we are the wrong choice, and in each one, who to pick instead.
A bespoke system pays off when the process behind it is genuinely bespoke. If the task is invoicing, project tracking or CRM in general, an off-the-shelf product is in place sooner and cheaper, and you are not paying for its upkeep.
Look at the off-the-shelf options first. If it turns out you would have to bend them in three places to fit your process, come back. Those three places are our ground.
We are not the cheapest, and we are not trying to be. A senior team costs more than a junior one, and the difference does not show up in the hourly rate, it shows up in how many times the work gets redone.
If you want a starter portion of AI, the Hungarian field sells bounded, published-price entry products: mini audits, workshops, monthly packages. These are real entry points, and you do not need to buy them from us: their current prices are on their own sites.
This is not snobbery. It is a question of scale. A website assistant can be stood up on a ready-made platform in half a day. If we built it, you would get the same thing for more money, and its upkeep would sit with us.
A packaged chatbot platform is faster and cheaper. We have something to do when the assistant has to reach into your own systems: record an order, check stock, issue an invoice. That is no longer a chatbot, it is a process.
What we build does not decide: it prepares, proposes and waits for someone to accept. That is not a limit of the technology, it is a design decision: a system that decides business questions on its own takes on responsibility it cannot carry.
If the goal is full automation with no human approval, that deserves a separate conversation before anything gets built. There are places where it is right; in most places where it is asked for, it is not.
We do not sell developer hours by the head. Most of our work runs in production at end-user companies, and the responsibility is ours there. A subcontractor role loses precisely the part that creates the value.
Joint delivery is worth discussing, though: you bring the client and the domain, we bring the system and the responsibility for it. That works. Rented capacity does not.
Someone who has had a system idea sitting around for years that nobody would take on. Too bespoke and too small for a software house, too specific for an off-the-shelf product, so it has been running in a spreadsheet, in email or in someone's head ever since.
If that sounds familiar, it is worth a conversation. Not because there is certainly a solution, but because an hour is enough to find out whether there is.
There are three of us. An enquiry that never starts is not a convenience, it is capacity: the time that goes into the systems already running. And a denial stays true for years: nothing to maintain, nothing to go stale, and no way to tell how long ago it was written.