A major global airline: premises Edges to hybrid media and five cloud media regions

The airline already ran its contact centres on a single Genesys Cloud CX organisation, with its core region in Tokyo and agents across Asia Pacific, Europe, North America and Australasia. Every call, though, still passed through premises Edges: 18 production Edge appliances in five data centres, connected to a private airline network and its session border controllers, plus Edges for test environments. Genesys has announced end of support for this hardware on 1 December 2026, and end of life on 1 June 2027, when remaining deployments stop working.

At the same time, the airline was moving its telephony from a broad mix of in-country carriers to a single global voice provider. That provider was a global systems integrator, which brought in QVCCS for the Genesys Cloud side. The airline's existing Genesys reseller, another global integrator, needed both kinds of expertise: global voice orchestration, and the hybridisation of the organisation so that it could use Global Media Fabric on AWS and leave the premises Edge problem behind.

QVCCS was the Genesys Cloud CX specialist for the incoming voice provider, working alongside the reseller's professional services and telecoms leads and the airline's own network, firewall, security and telecoms teams. Discovery began in June 2025 with technical workshops on network readiness, the Edge estate, endpoints and firewall rules, then a walkthrough of the hybridisation process against every premises Edge, and finally an agreed way to prove that the organisation was ready for the global voice migration that would follow.

1. Technical discovery: sites, data centres, Edges, networks and endpoints
2. Firewall and network requirements for hybrid and cloud media
3. High-level design for five media regions and five cloud sites
4. Hybridisation runbooks for SIT, UAT and production
5. Carrier SIP trunks, dial plans and outbound routes
6. Testing in each organisation, load testing and production go-live

The runbook followed Genesys's premises-to-hybrid media process, step by step, with an owner, a time and a check for each. BYOC Premises and BYOC Cloud were added to the organisation through the AppFoundry. The team then confirmed that Genesys had created the new phone trunks, the default location and cloud site, the Genesys Cloud Hybrid Media Group and its virtual Edges, and that live service was unchanged.

Each premises Edge was brought to the latest version and moved into the Hybrid Media Group, so premises and cloud Edges work as one cluster. Site interconnects were set on each Edge's network ports, a site link joined the premises core site to the cloud, and each Edge was restarted with call draining so no live call was cut. Test users on premises and cloud sites then proved calls in both directions between them.

The same runbook ran first in the SIT organisation, then in UAT, and only then in production, so every step had been rehearsed twice before it touched live service. A Genesys readiness case was opened in advance, given the size of the estate.

With the organisation hybrid, Global Media Fabric let each group of users and each trunk use the Genesys Cloud media region closest to them, while the core organisation stayed in Tokyo. Five cloud sites were built: one for Japan and Korea in Tokyo, one for South-East Asia in Singapore, one for EMEA in London, one for North America in N. Virginia and one for Australia and New Zealand in Sydney. Agents use WebRTC phones, so their voice travels to the nearest region rather than across the world.

Each site connects to the new global carrier's SIP trunks in its region, where resilient SBC clusters in separate locations host the airline's numbers. A dedicated trunk through the Asia Pacific SBC cluster to the Singapore region carries the numbers for six South-East Asian countries. Each site has national and international number plans and a default outbound route.

Outbound resilience is built into every site. Each site lists all five carrier trunks in order, starting with its own region and then the others, so if a region's trunk is interrupted, outbound calls continue through the next one. Agents in London, for example, fall back to the North American trunk, then Singapore, Tokyo and Sydney.

Inbound resilience works differently, and the design says so plainly. Calls to a region's numbers rely on the active/active SBCs within that region's carrier cluster, because inbound calls cannot fail over between carrier regions. Writing that down in the high-level design meant the airline approved the real behaviour, not an assumption.

Not every market could move at once. In some countries telephony does not support SIP and VoIP in the classic sense, so calls there still terminate on the airline's own SBCs and are routed up to Genesys Cloud directly. In one market, local regulation means agents keep using local premises Edges and a local carrier for domestic calls, with a second WebRTC phone for international calls, until a compliant replacement is agreed. That market is identified as a future media region.

The premises Edges were not switched off. Global telecoms migrations take far longer than a platform change, and every number move needs deliberate planning and extensive approval. So the Edges stayed in the inventory as spare media capacity, available while the global voice migration completes, rather than being retired early. That spare capacity was a deliberate and sensible choice.

The airline now has one hybrid Genesys Cloud CX organisation. Its core is in Tokyo, with media in Singapore, London, N. Virginia and Sydney, cloud sites for each region and carrier trunks with cross-region outbound failover. Premises Edges and cloud Edges sit together in the Hybrid Media Group, so calls route between every contact centre site and any trunk, wherever it terminates.

Agents connect to Genesys Cloud over the internet and use WebRTC phones. The private airline network is now needed only to reach the remaining premises Edges, and firewall rules for Genesys Cloud media were set out region by region and reviewed by the airline's firewall team.

• One hybrid organisation, core region Tokyo
• Global Media Fabric in five regions
• Five cloud sites, each with all five carrier trunks in failover order
• 18 premises Edges in the Hybrid Media Group, kept as spare capacity
• SIT, UAT and production organisations hybridised in turn

Testing followed the same three stages, with written test documents for SIT, UAT and production. They covered calling between cloud and premises users, inbound and outbound calls on test numbers through each regional trunk, the existing premises trunks, and failover. Production went hybrid on 18 October 2025.

The airline's IT rules also required load and performance testing. The carrier ran spike tests of 300 concurrent calls at 10 calls a second, and soak tests of 100 concurrent calls for an hour, about 1,000 calls each, in North America, the UK and Australia, and every call connected. In truth, such tests say little about Genesys Cloud CX: it is an elastic, self-healing microservices platform on AWS that scales automatically, so one customer's test traffic reveals almost nothing about its capacity. The requirement came from rules written for legacy on-premises systems, so the tests were run as the airline required.

Number testing found one real fault: calls from mobiles to some international numbers failed as unregistered. The cause was the global carrier's SIP backbone dial plans, which had misrouted those numbers to a local network point of presence. The carrier corrected the dial plans, and it was resolved quickly.

The airline's Genesys Cloud CX organisation is hybrid and ready for the end of the premises Edge hardware, well over a year before end of support. Its agents use cloud media close to where they work, with outbound calls able to fail over across five carrier regions.

The hybridisation also cleared the way for Part 2: the global voice migration from many in-market carriers to a single global voice provider. That ran in parallel, and QVCCS supported it, especially the SIP trunks and initial testing, but it continued long after the hybridisation was complete, as global telecoms migrations do.

• Production hybrid on 18 October 2025
• More than 400 agents on cloud media in five regions
• Outbound failover across all five carrier regions
• Premises Edges kept as spare capacity during the carrier migration
• Every load test call connected