Grow from fifteen people to eighty and the technology usually stays the technology you had at twenty. One administrator, perhaps an apprentice alongside him, a file share that everybody can write to, and a business system whose restore procedure exists in exactly one head. Then a large account posts over a supplier assessment shaped by NIS2, asking for evidence about second factors, tested restores and how incidents get reported. We extend that small team over the network, across the ground it has neither the hours nor the tooling to reach.
At this scale, whether Monday morning goes well no longer depends on a salesperson's laptop. It depends on machines, directories, permissions and stored data. Your own people keep the business knowledge; the plumbing becomes our problem.
Windows and Linux workloads, whether they sit in your own plant room, at OVHcloud or Scaleway, or on Azure. Updates go out inside maintenance windows you have agreed to, and each change leaves a dated line in a log somebody can read afterwards.
Your directory gets reworked around membership per department, replacing access that was granted three years ago to unblock somebody in a hurry and never reviewed since. Someone joining is productive by lunchtime; someone leaving loses everything before the afternoon is over.
Workloads, cloud mailboxes and the business database copied on a schedule, with one set held well away from the building. On the rhythm set in the agreement we bring a randomly chosen system back and record how long that took. Until somebody has restored it, a copy is only a theory.
Firewalls, switches and access points configured and watched from our side. Visitors, cameras and shop-floor equipment are separated from the desks that handle payroll, and a mobile standby circuit carries traffic on the afternoon the fibre goes quiet.
We sit with you over the security questionnaires your accounts send and assemble what they want to see: the state of second-factor sign-in, a written copy policy, the route an incident takes when it is reported, a record of who holds elevated rights.
On the agreed rhythm, a statement pulled from our ticketing: what came in, how fast it moved, where patching and copies stand, and the risks we can see taking shape. Enough substance to build next year's budget rather than defend last year's spending.
The decision is seldom either-or at this scale. Three arrangements turn up again and again, each with something it does well and something it quietly leaves undone.
Arrangement 1
Handled entirely in-house
A single administrator owns the lot, from toner cartridges up to the firewall rules.
Arrangement 2
Duties divided
Your administrator keeps the business system and the users; infrastructure and protection come to us.
Arrangement 3
Handed over completely
The whole of it under a single agreement, with one named person answering for it.
Possibly not directly. The directive and the French law transposing it may well leave your own company outside their scope, while catching the customer who wrote to you, who must then look at the firms in their supply chain. That is where the questions about second factors, copies and incident reporting come from. Whether the text binds you is a matter for a lawyer. Our part is answering the technical items accurately and, where a control is genuinely absent, building it before the same account asks a second time.
The first pass simply records who holds what, and separately who exercises it. The distance between those two lists is always worth seeing. Definitions are then agreed with the heads of each department, and people move across in small batches, with a route back available within minutes. How long it takes depends on your size, and every batch is scheduled away from the days when the business cannot afford surprises.
Not at all. Nobody else knows your machines, your habits and your colleagues the way he does, so he stays where he is. What leaves his desk is the repetitive work, the alert watching, and the narrow specialities such as firewall rules, device management and restore testing. That frees him for the business system and the projects, and incidentally makes a fortnight away possible without the work laptop.
It starts with a review conducted over the network. Several early weeks go into cataloguing workloads, services, dependencies, elevated accounts and carrier agreements, then writing down what was never written down. That file belongs to you throughout, including on the day you decide somebody else should hold it.
Your publisher is not being replaced, whether that is Sage, Cegid, Divalto, Odoo or something written for you years ago. We own the layer beneath the application: the host, the database, the response times, the copies, the remote access, the update windows. When the publisher does connect, it meets a documented environment, and you stop refereeing between two accounts of the same failure.
Send over your headcount, how many workloads you run and what sits on them. We look at the existing setup and return a figure with a list of what would actually be done.
Message received
An answer follows inside one working day. Report an outage that is stopping people working and it moves ahead of everything else.
Nothing here under that name. Check the spelling, or simply choose the nearest large city instead. Since every engagement runs remotely, whichever you pick changes nothing about what we do for you.
