Managed Run takes over the round-the-clock run of your infrastructure with agent teams. You get service levels you can check against your own tools, and a named person who can stop the whole fleet at any hour.
Every seat you cover around the clock takes several people once leave, training and the working year are counted, plus the out-of-hours premium in your contracts. Agents cover the night at no extra cost.
What still needs a person at night is a decision nobody wrote a rule for, or a crisis. That is a much smaller rota than the one you staff today.
Before any agent runs, we measure your run as it is today: volumes, delays, recurrence, callouts. You can recompute every figure. Without a before, nobody can argue about the after.
Every figure shows the period it covers, the system it came from, and how much of that period the system actually saw.
Each kind of action has an error budget. Spend it and that action goes back to supervised at once, with no meeting. Going back up is a person's signed decision.
No agent approves a change or closes an incident. Every escalation says what happened, what the person must do, and by when.
A named person can halt one agent or the whole fleet, at any hour, and the stop is recorded.
An agent changes your estate only under a change your people approved, or a standard change they pre-authorised, inside its declared bounds. Anything else is refused and counted.
Your configuration database is checked against what your estate really contains. Where it is incomplete or stale, agents there stay supervised, however well they perform.
An agent acts on one object at a time, says what it changed, and checks the result in your own tools before it calls the job done.
Per kind of action, never in one switch. Each one climbs on its own record.
Agents connect to your tools and read. Nothing acts. You get a measured baseline.
Agents say what they would do on your real tickets. Nothing is executed. Their calls are compared with your engineers'.
The agent prepares the action. A named person approves it in one click.
Granted by a person, on evidence, one kind of action at a time. A spent error budget sends it back to supervised.
Your run today, frozen and signed. Your inventory reconciled with your CMDB.
Event triage, low-priority incidents, standard requests, backups, patching.
Servers, virtualisation, cloud, storage, on runbooks your team wrote.
Databases, network security, continuity: watched and advised, not touched.
One team per tower of your run: Supervision Desk, Network Operations, Database Operations, Cloud Operations, Server & Virtualisation Operations and Request Fulfilment, with Incident Neutralisation, Change Guard, Fix Squad, Recovery Proof and Identity Hygiene. They work in the tools you already run and your service-management tool stays the record. Coming in, tower by tower: Run Takeover. Leaving: Reversibility Dossier. See every offer.