A 25-person professional services firm in Bristol

The problem

The firm wanted to leave a local provider whose helpdesk was a ticket queue and whose answer to most problems was a visit charged by the hour. An ageing Active Directory server and a file server in the office that everyone depended on, hybrid staff, a Microsoft 365 tenant that had been migrated in a hurry and never finished, and CCTV and network kit of various ages that nobody owned.

Every one of those was a single point of failure, and a Brighton provider asking for the work had to show it could remove them rather than promise to drive to Bristol when they failed.

What we did

The single points of failure came out at the start of the service as one planned project, not a slow drift. Identities moved from the on-premises domain to Entra ID, files from the file server to SharePoint with permissions designed rather than copied, and both old servers were decommissioned and gone. The network was replaced with UniFi, with its configuration backed up to the cloud, and the old CCTV replaced with UniFi Protect cameras on the same system, so one console covers the network and the cameras.

Every device went into Intune, so any of them can be rebuilt from the box without a visit. Conditional Access, a backup covering SharePoint and OneDrive, and a written disaster recovery plan that assumed the office itself might be unavailable.

Since that project, eighteen months ago, nobody has needed to be in Bristol for anything routine. Support is remote and fixed within minutes of a call.

What changed

The exception proved the set-up. A ceiling leak over the comms cabinet took out the UniFi network equipment. The disaster recovery plan was initiated: with identities in Entra ID and files in SharePoint, every member of staff worked from home with no loss of access to anything. That is what the plan is for, and it is why nobody needed us in the building within the hour.

Replacement kit was ordered that day and an engineer was on site with it at 9am the next morning. The new gateway and switches were adopted and restored to the previous configuration from the cloud backup rather than rebuilt by hand, and because there was no server in the building the leak had nothing else to ruin. The office was back online within a few hours of the engineer arriving, with limited impact on the business.

The leak exposed one risk that had been accepted: the comms cabinet's position under that ceiling. A follow-up job to recable the office fully and move the comms equipment to a different location is booked as planned works rather than another emergency.