IT infrastructure
Networks, servers and the daily operations around them, for companies whose IT grew by accretion rather than by design. Run by people who have done it for a decade, and documented so you are never locked to us.
What this looks like in practice
Work we have actually done, described as work rather than as capability.
- Networks and servers past consumer kit
- The point where the office router and a spare desktop stop being enough, and downtime starts costing more than the replacement would.
- Backups that have been restored at least once
- A backup nobody has restored is a belief, not a backup. We test the restore and write down how long it took.
- Monitoring that reaches a person
- Alerts that wake someone when it matters and stay quiet when it does not. A dashboard nobody opens is not monitoring.
- Documented handover
- Credentials, diagrams and runbooks written so another team could take over without talking to us. That is the point.
Constraints we work with
The limits that shape this kind of project. Naming them costs us nothing and tells you more than a list of technologies would.
- Hardware that will not be replaced this year
- The budget exists or it does not. We work with what is on the floor and say plainly which risks that leaves open.
- The work happens outside business hours
- Switch replacements and migrations land at night or on a weekend. Planning that honestly is part of the estimate, not an afterthought.
- Requirements written before the cloud existed
- Some rules predate the options they forbid. We keep to them and note where they cost you money, so the choice is yours to make.
- You are ten thousand kilometres from us
- Anything needing hands on hardware needs someone local. We plan for that from the start rather than discovering it at the outage.
Where this shows up in our work
No public case study here yet. This work is mostly ongoing operations rather than a project with a start and an end, and the clients it belongs to have not been asked whether we may write about it. Rather than invent an example, we would rather answer questions directly — ask us what we run, and we will tell you.
Further reading
- Declare maintenance jobs as data, not as codeA maintenance tool that can run any command is one nobody should trust with admin rights. What changes when the job is declarative data instead of code.
- Why Windows housekeeping only works right after a rebootThe files that most need cleaning are the ones Windows is holding open. Why the minute after a reboot is the only dependable window, and how to use it.
- OfflineTasksAn open-source Windows maintenance tool that runs cleanup jobs at the one moment they reliably work: right after reboot, before anything can lock a file.
Infrastructure held together by one person who knows?
Tell us what you run and who currently keeps it alive. The first useful output is usually a written picture of what exists — which is often the thing nobody has.
Start a projectWe reply within 1 business day. No sales calls unless you ask for one.