Your systems are yours. We work by rules you can check.

Every access starts as read-only. We look at configurations, logs and state, but change nothing. Reviews, assessments and diagnoses are done without a single change on your systems. The right to change is granted separately, per specific change.

Before each change we describe what we’ll change, why, and what could go wrong. We make the change only once you approve it. The approval covers that change — not the next one and not the whole project. Silence is not approval.

Before we change a configuration, file or database, we make a dated copy of the previous state. No exceptions, even when the change is “just one line”.

Each change comes with how to return to the previous state and how long that takes. If we don’t know how to roll it back, we don’t do it.

We don’t assume the server matches the documentation or our copy. Before a change we diff against the server’s state and verify that what we deploy is exactly what we prepared.

We don’t make changes before or during an event that must not stop for you — month-end closing, a delivery, a live event, a render deadline. The change window is agreed with you, in advance.

If other systems live on the same server or account, we leave them alone — and after every change we verify they still work.

We don’t read the contents of mailboxes, documents or databases when that isn’t part of the agreed job. We work on configuration, logs and metadata.

After every job a record remains: what was there, what changed, where the backup is, how to roll back. The documentation is yours — and with it you can continue with us or with anyone else.

We don’t sell or resell software licenses or “catalog” hardware. We do setup, tuning, networking, administration and support of your equipment and legally acquired software — nothing else. Licenses stay yours, and your responsibility.

If the cause wasn’t found, we write down what was checked and ruled out. If something isn’t our field, we say so and recommend who to call.