Independent, and answerable for it.
8lax is a software studio in Ukraine, founded and run by Oleksandr Lytvyn. Independent: not part of a group, not an agency network, and not answering to anyone whose money would decide what gets built.
Five products across four unrelated markets, all of them ours. No client projects, no outsourcing: everything here is built to be sold to anyone who has the problem, which is a slower way to make money and the only way to own what you make.
Why the work looks like this
Software is sold on words that nobody checks. Immutable. Bank grade. AI-powered. Fully automated. Look closely and the immutable log is a database table the vendor can edit, the automation is a CSV export, and the AI is a prompt with no way to tell when it guessed.
The alternative is not better adjectives. It is building so the claim can be tested by the person who doubts it: evidence packs that verify on a stranger's laptop with our software nowhere in sight, and an estimate that arrives as a range with its assumptions printed beside it rather than as a number you have to trust.
What independence changes
You reach the people who wrote the code, and the founder reads it. No support tier, no ticket routing, no "I'll pass this to engineering", which is the part of buying software from a large vendor that nobody misses.
The obligation that comes with it is that the products have to be built to survive us, whatever happens to us: read-only by design, data in your own tenancy, and an export format documented well enough to outlive the app. A vendor you can walk away from is the only kind worth trusting.
Where the judgement comes from
A studio's opinions are only worth as much as the work they came out of. Here is the work.
Fifteen years of backend engineering, most of it for businesses where a mistake is expensive: fiscal software that has to satisfy a tax authority, risk valuation for the German construction industry, investment and payment platforms, and high-availability retail systems carrying real traffic. Before 8lax, tech lead and team lead, running delivery for teams of four to eight engineers, QA and analysts, which means sprint planning, estimating work, and deciding who was free to take it.
That last clause is why ScopeBench exists. It automates a job done by hand for years, on Sunday evenings, badly.
The compliance work has the same origin. Building software that has to hold up when a regulator asks is a discipline rather than a feature, and it is the one this studio keeps coming back to. It is where the habit of writing "tamper-evident, not immutable" was learned, because somebody eventually reads the small print out loud.
AI is not new here either. The rule about checkable output came from shipping model-backed features into a production retail system, where the first thing you learn is how confidently a model gets it wrong.
How something becomes a product here
Four markets that have nothing to do with each other is either a studio with a method or a studio with attention problems. Here is the filter.
A dated, specific problem
An auditor asking who approved access in March. A rota that has to hold together tomorrow morning. A client email that needs an answer this week. Vague problems produce vague products, and vague products never find the person who would have paid.
Checked against the market first
Before the build, we go and find who is already there and what they charge. It has already stopped one product from being built the way it was planned, which cost a fortnight of research instead of a year of work.
Small enough to ship
A product that is honest about its edges can be built and put in front of someone in weeks. One that needs a tender process and six integrations before anybody can try it cannot, so we do not start those, however large the market slide says it is.
Where this is going
Stated plainly so you can hold us to it.
TopBin9 is live and gets the work that a live product gets. Attestly is being prepared for the Atlassian Marketplace. CarePlanner is in closed beta and needs real carers using it before it goes further. ScopeBench is built and waiting on its first outside users. LiCly has a working consumer build and a business case still in research.
What happens after each of those is shaped by what the people using them report, not by a roadmap written in advance. If you want to know whether something specific is planned, ask: the answer is either yes with a date or no, and there is no third answer worth giving you.
The direction we want next is public and civic work: services people are obliged to use rather than choose, where the same repeated hours cost a great deal more because they are paid for by everybody. Nothing is built there yet, so it is an intention and not a line of business, and this paragraph will either name a product or disappear.
The studio is growing into a company. What will not change with the size of it is which claims we are prepared to put in writing.