The questions people ask before signing anything.

We do not publish rates, because an honest number depends on what discovery finds. Uncertain work starts with a bounded, paid discovery phase, usually one to two weeks, with the scope and price agreed before it begins. Production is priced after that, once the data, integrations, and quality bar are known rather than guessed.

It is set in the contract before work starts, and both models are used. Some engagements assign the rights in the delivered code to you. Others licence it to you while we retain the underlying IP. In both cases we keep our own pre-existing and general-purpose components and licence those to you as part of the delivery, and the contract lists exactly which ones. Nothing about ownership is left to be discovered at handover.

In our standard engagements the contract sets a warranty period during which we fix defects in what we built. That covers defects rather than new features, which are quoted separately. After the warranty, support is paid and runs in business hours. If the system needs someone to keep it alive, meaning the cloud environment, deployments, monitoring, and dependency and security updates, that runs on a monthly fee covering the system staying up rather than growing. We do not offer nights, weekends, or 24/7 on-call, and we would rather tell you that now than during an incident. Warranty length, support hours, and the monthly fee are set in the signed agreement, not here.

The two founders, directly. There is no sales layer standing in front of an unnamed delivery team. You meet the people who would design and build your system on the first call, and they stay on the work through implementation and handover. Where a project needs a specialist we do not have, we bring one in under our own contract with you, so you deal with us and we stay accountable for the result on the terms we signed.

We aim to confirm a fit call within two Romanian business days of your enquiry. Once that call has happened and we have the access and data the work depends on, discovery normally starts within five business days, and the proposal follows discovery rather than preceding it. Those are the intervals we work to, not guaranteed deadlines, and the honest answer for your specific case comes on the call.

You pay for the phase in progress, and beyond that only costs already committed on your behalf, which the contract lists before the phase starts. Everything that phase produced, meaning the evaluation, the findings, the documentation, and the code as it stands, is handed over on the terms agreed before work began. That is the point of staging the work: each phase is small enough that stopping is a reasonable decision rather than a write-off, and we would rather you stop at a gate than continue out of momentum.

Yours, under your own billing. We work inside it with scoped access rather than owning it, so you are never renting your own infrastructure back from us, and you can revoke our access at any time. The contract says what that means for work in progress, so neither side is guessing. Where a build and test environment is needed we use our own, which keeps work in progress out of your production account. It runs on anonymised or synthetic data by default, and if real data has to be used that is covered by a data-processing agreement signed before anything is copied.

It keeps running in your own subscription rather than ours. What changes is maintenance, not hosting, which is why the contract sets the exit and handover terms up front. The build is designed for that from the start: documentation, a runbook written for whoever is on call, no tooling that only we understand, and access and IP terms agreed before work begins rather than negotiated when somebody wants to leave.

We prefer to own the delivery end to end, because split accountability is where systems fall between two teams and nobody is responsible for the seam. That does not shut your engineers out: they review the work, take the handover, and run it afterwards. What we will not do is accept responsibility for a system where the parts most likely to fail are somebody else's to fix.

Yes. We do not sell speculative foundation-model training, standalone cybersecurity, brochure websites, or round-the-clock founder support, and we say so on the first call rather than after a proposal. Beyond that we look at the work itself rather than the sector label. Before we quote we check that we may lawfully do it: sanctions and export-control screening, and whether the intended use falls into a category EU law prohibits or treats as high-risk. Sectors carrying licensing or regulatory duties are not excluded, and those duties go into the contract before anything is built. Where the intended use is one EU law does not permit, no contract fixes that and we decline.