01
"The people who built my system disappeared"
They changed jobs, shut down or stopped replying. The software is still running and nobody knows how it is built inside, so nobody dares touch it.
Evolutionary software maintenance
We mean maintaining the software your company uses: fixing it when it fails, updating it when the environment changes and evolving it when the business asks for something new. This is not software for managing the maintenance of machines or equipment: it is maintenance of the system, done by people who read the code.
The breaking point
These three sentences are the ones we hear most on the first call. They are not rare cases: it is what happens when a development is delivered and nobody stays in charge.
01
They changed jobs, shut down or stopped replying. The software is still running and nobody knows how it is built inside, so nobody dares touch it.
02
Updates are not optional —there is security at stake— but every time they are applied something stops working. In the end they stop being applied, and the problem only gets bigger.
03
The business changed: a different process, a new tax, another sales channel. And every adjustment comes in as if it were a project from scratch, with its own wait and its own negotiation.
What we do
Software maintenance has three faces and the third is the one almost nobody covers. Fixing what breaks is the bare minimum; what keeps a system alive is being able to change when the business changes.
We take the error, look for its cause in the code and fix it at the root. When the same fault comes back twice, the problem was never the fault: it was the cause nobody looked for.
Language versions, libraries, plugins and certificates kept current, applied with a test before they reach production. That way updating stops being the day something falls over.
Adjustments and new features on top of what already exists: one more field, another report, a change of process, an integration that was not needed before. The system follows the business instead of holding it back.
So there is no confusion
The two are searched for with almost identical words and have nothing to do with each other. We spell it out here so you do not waste time if you arrived looking for the other one.
You have a system, a web platform or an application your company already uses, and you need someone to fix it, update it and evolve it. That is what we do.
If you are looking for a program to schedule the maintenance of machines, vehicles or equipment —what the market calls a CMMS— this is not that kind of product. Needing one built to measure is a different matter: that would be a development project.
We do not do industrial maintenance, or maintenance of computers, printers or physical networks. Our work is on the code and the data, not on the hardware.
How it works
We do not publish response times on this page: they depend on how critical your system is and on the scope we agree. What we can tell you up front is the mechanism, which is the same for everyone.
Everything that breaks and everything you ask for comes in through the same place and gets recorded. Nothing lives only in a chat or in somebody's memory.
A person who knows your system answers for what comes in. You do not start from scratch explaining the context on every request.
What stops the operation and what can wait are classified by a criterion we define with you, not by order of arrival.
Response times, the coverage window and what is in and out of scope are agreed in the proposal. That is where the numbers belong, and they are the ones for your case.
Scope
Taking over the maintenance of something built by someone else starts with understanding it. These four points are what we review before committing to anything.
Either the repository is handed over or we rebuild it from the server. Without being able to read the code there is no maintenance, only attempts.
A system on very old versions sometimes needs bringing up to date before it can be maintained normally. That is spotted and said from the start.
A stable operation needs little evolutionary work. One that changes process every quarter needs a lot, and that changes how it is contracted.
An informational site and a system with people working inside it eight hours a day do not call for the same coverage or the same commitment.
The first month with someone else's system is about getting to know it: understanding the code, the data and the integrations before promising anything. Book a free assessment.
Where it fits
This page is about one thing only: software that already exists and has to be kept running and growing. The operations relationship is wider than that, and includes things this page does not explain.
Corrective work, updates and evolution on a system in production, including when another supplier built it and it has to be taken over.
The support desk, monthly digital operations, reporting, technical direction and the contracting models live in application support and maintenance, the hub this page belongs to.
And if what has to be kept running is also the server the system lives on, that's managed IT services. If the system still has to be built, start with software development.
Fit
If it is a single error and you do not want a relationship afterwards, it can be done, but it is not what we do best. The value here appears when someone stays and gets to know the system.
Sometimes the code is so far behind that keeping it alive costs more than building it again. When we see that, we say it: we would rather lose the maintenance work than charge you for propping up something with no future.
Questions
Yes, and it is most of our cases. We need access to the code and to the server. The first month is about getting to know it: we understand how it is built, what data it handles and what it integrates with, and out of that comes what can be kept as is and what should be fixed first.
Corrective work returns the system to the state it was supposed to be in: something broke and it gets fixed. Evolutionary work takes it to a new state because the business changed: a different process, a report that did not exist before, a new integration. Most contracts need both.
It depends on how critical the system is and on the priority of the case, and that is why we do not publish a figure here: it would be made up. The mechanism is fixed —single channel, assigned owner and agreed priorities— and the committed response time is put in writing in the proposal.
It is a common scenario when the previous supplier disappeared. The code can often be recovered from the server the system runs on. If it cannot, it has to be said plainly: without code there is no real maintenance, and the conversation becomes a different one.
Yes. Core and plugin updates tested beforehand, fixing what breaks on update, and building whatever the site needs. It is one of the most common "it breaks every time I update" cases.
Monthly, on a block of hours or as an improvement project, depending on how much your business changes and how much you need. The models are explained in the operations hub and which one fits is decided in the diagnostic.
Next step
With that we tell you whether it is worth maintaining, bringing up to date first, or rebuilding. And if the answer is rebuilding, we say that too. No strings attached.
Prefer to write to us? Tell us what you need.