Handover
Every job ends with a handover, and every handover ends with a pack. Drawings, schedules, certificates, datasheets, a list of passwords the client was told to change. It gets accepted, filed on a shared drive, and opened again only when something has already gone wrong.
On a modern building that is only half the problem. The fire alarm, the cameras, the access control, the network and the AV each arrive from a different installer, and each comes with its own portal showing only its own kit. The facilities manager ends up with five logins, five versions of the floor plan, and no single answer to the question they ask every morning: is everything working?
A building manager has four questions about the systems you fitted.
Is it working right now? Not on the day it was commissioned. Now, before the occupants notice.
Where is it? The camera that has dropped off the network is somewhere on level three. The pack has its model number, not its ceiling tile.
Who fitted it? And how do they report a fault without a paragraph of directions you then have to decode?
What have we got? For the insurer, the budget and the next refit, as it is today rather than as it was when the binder was printed.
None of that is in the pack in a form anyone can use at eight in the morning with a complaint already in.
If your engineers pinned every device on the drawing as it went in, you are holding the most useful document the client will ever get: the real plan, with every device where it was actually fitted. It just lives in your system, not theirs.
Graftix Live hands it over. The client signs in at graftix.live and sees your drawings with the devices on them, with live status on the networked kit. There is nothing for them to install; it runs in the browser and can sit on a phone’s home screen. A native Graftix Live app for iPhone and Android is on the way.
A globe of their sites, then a map of the campus. Every site with a health colour on its pin. On a campus, one pin per building, so the estates team can see which block has the problem before they open anything.
Plans grouped by installer and by system. Fire alarm, CCTV, access, data: each installer’s drawings, labelled by system, building by building. Shares from different installers for the same building land in the same place.
The real drawing, with live status. Every monitored device where it was fitted, refreshing every 30 seconds. Tap one for its status history, uptime over 24 hours and seven days, make, model and serial where recorded, and who installed it.
Events and an asset register. A timeline of what went offline and came back, and a register of every shared device that exports to CSV.
Messages and fault reports. One conversation per installer per project. A fault can be pinned to the exact spot on the plan, so it reaches you with a location rather than a description.
Alerts. An email when a device has been offline for longer than they choose. One email per outage, not one per blip.
Plain-English answers. Ask Dave answers questions like “what is offline right now?” from their own sites. If they already use Claude or another assistant that speaks MCP, they can connect it to their sites directly.
Nothing is shared until you share it. Client Share sits on the project’s Live page in Graftix, and you decide which plans go across, each labelled with its system; whether live devices are included; whether IP addresses are shown; and who pays, you or the client. Devices on plans you leave out stay private.
Add a message and send it. The client gets their own organisation and invites their own team. You can pause or revoke a share at any point.
When the client reports a fault, you see it in Graftix with its position on the plan, and one click turns it into a snag task at that spot. The report becomes work with a location and a status, not an email somebody has to retype.
Your tasks, timesheets, purchasing, qualifications, RAMS and team. None of it. Client organisations are refused by the engineering side of Graftix on the server, not just hidden in the interface.
graftix.live also holds no copy of your data. It has no database of its own; it is a front door to Graftix that only lets through the calls the portal needs. Sign-in stays on graftix.live, and the session never reaches the browser’s JavaScript.
Status comes from the Graftix Live agent, a small service on any Linux box, Raspberry Pi, Docker host or Windows machine that can reach the kit, talking out over HTTPS with no inbound firewall rules. Or straight from ExtremeCloud IQ, if the network is Extreme. Cameras, access points, switches, servers: anything with an IP address.
Worth being clear about the limits. A device shows live status once it is placed in Graftix, on a pin or a task. Graftix Live does not read fire alarm panels: the fire drawings can go across with everything else, but the panel stays the authority on the fire system’s own state.
£10 + VAT per monitored device, per year, licensed per project. Every project gets a 30-day trial. Either side can pay, and licences can be bought online or on invoice. A lapsed licence only changes what the client sees; monitoring inside Graftix carries on.
Pick one building where the devices are already pinned and the agent or XIQ is already reporting. Share one system with the facilities manager, IP addresses hidden, and see what they do with it over a fortnight. Whether they report the next fault through it, or still ring, tells you most of what you need to know.
Try it on a real job
Graftix puts every task on the actual drawing, then hangs the materials, certificates and sign-offs off that pin. Seven days, every feature, no card.
Read-only live environment · No sign-up required