Three bottlenecks in multi-site work
Building one site is easy. Copying it eight hundred times without errors, without closing the doors, and in a way that can be signed off — that is the work.
Boxes stacked, configs drifting
A traditional branch needs five boxes — router, firewall, switch, APs, PBX — each configured on its own. The more sites, the worse the configuration drift.
Ageing voice systems
Legacy PBX lines are worn, mostly analogue handsets, they cannot reach the head-office directory or the contact center, and they are expensive to keep.
Invisible and unmanageable
Hundreds of scattered sites with no single view. Faults arrive by phone and diagnosis means driving there.
The common answer: move the complexity off each site and into the cloud.
Architecture: a lean site, controlled centrally
One gateway, four jobs
A converged gateway combines routing, next-generation firewall, PoE switching and IPSec VPN, with APs and IP phones behind it. Five boxes become two.
Control moves to the cloud
A cloud controller manages every AP in the province, pushing configuration in seconds and upgrading firmware in bulk; cloud IP PBX makes inter-site calls free and ties into head-office systems; the audit and authentication platform meets log-retention and real-name requirements.
All traffic backhauled encrypted
Audit and authentication data return to the cloud over IPSec VPN tunnels, end-to-end encrypted, meeting financial-sector data requirements.
One view of the province
Device state, signal quality, throughput and alarms for every site in one place, filterable by district or batch.
Delivery: a standard process
Stock ahead of time
Sourcing starts the day the contract is won: material is packed per site by AP count and labeled with the site name, so a crew installs the moment it arrives.
Pre-loaded templates
Every gateway and AP ships with a base template; on site you only cable and power up, and the device pulls its own configuration from the cloud.
Nights, in parallel by district
Work runs from 20:00 to 06:00, with fifteen crews moving district by district in parallel. The doors open as usual.
Digital acceptance
No paper spot checks. The platform generates a heat map and a device health report per site, so acceptance happens at completion and passes on the spot.
The result: 30 days, 800+ sites
A branch refresh for a leading bank across Guangdong. The estate was old, voice was analogue and nothing was centrally managed; the requirement was no interruption to business, no site left overnight.
Acceptance produced 800 individual heat maps and device health reports; management can see the live state of any branch from their desks.
Where it fits: and where it does not
| Fits | Does not fit |
|---|---|
| Twenty sites or more, all much the same shape | A handful of sites, each with a completely different network |
| Head office needs one view of every store and site | Each site has its own IT team and will not hand over control |
| Fixed trading hours that allow night work | 24/7 operation with no window for downtime at all |
| Voice over IP or multi-site networking is needed | You only want to buy equipment, not manage it centrally |
| Clear audit requirements (finance, government, healthcare) | The budget covers hardware only, not installation or operations |
Stating the boundary is more useful than claiming everything. If the right-hand column applies, we say so up front instead of taking the work and improvising.
How we start
Three representative sites as a pilot
The cheapest way to prove the architecture, the build process and the acceptance criteria — then the volume conversation runs on measured data.
Templates and batches
Equipment list, configuration templates, packing rules, working hours and district sequence become a work instruction anyone can repeat.
Batch by batch, accepted batch by batch
Delivered by district; each batch ends with an acceptance report and a review, and the fixes land before the next batch.