Project
A defined scope, a fixed price and a delivery date. Suited to a first version, a replacement for something that has stopped scaling, or a well-understood piece of work.
Vorteks is a software engineering studio. We work with a small number of clients at a time on systems that have to be correct, maintainable, and understood by the people who inherit them.
What we are engaged to do, and what you receive at the end of it.
End-to-end development of web applications and SaaS products — data model, API, interface, deployment and monitoring. Built as one system by one team, so responsibility for how the parts fit together sits in a single place.
DeliverableA deployed system, its documentation, and a handover.
Language-model features specified, evaluated and costed before they ship. Retrieval, structured extraction and agent workflows, together with the evaluation harness that shows whether they still work after the next change.
DeliverableA measured system with an evaluation suite and cost model.
An assessment of a system you already have or are about to build: data model, service boundaries, failure modes, operational cost. Useful before a rewrite, a funding round, or a decision that is expensive to reverse.
DeliverableA written report with prioritised findings, and a working session.
Continued work on a system we built or inherited: new features, maintenance, dependency and security updates, and steady reduction of the components that generate incidents.
DeliverableA monthly retainer with agreed availability and response times.
The same five stages on every project, scaled to its size.
One to two weeks. We establish what is being built, what is explicitly excluded, and which questions are still open. Fixed deliverable, fixed fee, and no obligation to continue afterwards.
Data model, service boundaries, infrastructure and deployment plan — written down and agreed before implementation starts. These are the decisions that are cheap now and costly to revisit later.
Weekly increments on a staging environment. Each increment is deployable and reviewable, which keeps the gap between an assumption and its correction to about a week.
Production deployment with monitoring, backups, error reporting and an operations runbook written so your team can use it without involving us.
A defined support period after release. Performance work is done where measurement shows it is needed, and not before.
Three ways to work with us. Rates on request.
A defined scope, a fixed price and a delivery date. Suited to a first version, a replacement for something that has stopped scaling, or a well-understood piece of work.
Architecture review, technical due diligence, or a second opinion before a decision is made. Often the cheaper answer, and sometimes the conclusion is that no build is required.
Continuing engineering on a live system, with agreed availability, response times and a monthly allocation of hours. Available after a project, or for a system we did not build.
How the practice is set up, and what follows from it.
Vorteks is small by choice. The people you speak to are the people who write the code, which keeps the distance between a decision and its consequences short.
We hold two or three projects at a time. That limit is the reason we occasionally decline work that would otherwise be a good fit, and the reason the work we do accept gets attention rather than a queue position.
Our preferences are conventional: established technology over recent technology, tests where they earn their place, boring infrastructure, and systems that can be operated by someone who did not build them. Where we disagree with a direction, we say so before the work starts.
Enquiries are read by the people who would do the work.
Describe the work: what the system does, who uses it, when it is needed, and the budget range you are working within. Budget is not a filter — it tells us which of the three engagement types is realistic.
We reply within two working days, including when the answer is that we are not the right fit.
hello@vorteks.studio