Cloud replatforming and 24/7 operations for a logistics group

Our client is a logistics group operating warehousing and line-haul transport across Belgium and the Netherlands. Daily operations depend on a dispatch and track-and-trace platform: planners assign loads, drivers report status from mobile devices, and EDI links exchange orders and delivery confirmations with retail partners around the clock.

The platform had grown over a decade on servers in a single datacentre. A small internal IT team kept it running alongside office IT, without a formal on-call rotation or a documented recovery procedure.

Availability had become a commercial risk. Outages during evening loading windows stopped dispatch for entire depots, and incidents at night waited for the morning shift. The servers had been configured by hand over years, so nobody could rebuild them from a definition. Backups existed, but a full restore had never been tested.

Management faced a choice: build and staff a 24/7 operations capability internally, or transfer the platform — and the responsibility for keeping it up — to a partner under contract. They chose the second option.

We started with a four-week discovery and audit: dependency mapping, load profiling, and a restore test that confirmed the backup risk was real. We then designed an AWS target architecture in eu-central-1 (Frankfurt) to keep operational data inside the EU. The .NET dispatch backend was containerised and moved to ECS Fargate behind an Application Load Balancer; the SQL Server database moved to RDS with Multi-AZ failover and point-in-time recovery; the EDI gateway was rebuilt as a queue-based service on SQS, so partner traffic survives downstream restarts instead of being lost.

The entire environment is defined in Terraform and deployed through GitLab CI. Blue-green deployments replaced the quarterly release freeze: each release goes to a parallel target group, is verified against synthetic dispatch transactions, then switched over — or rolled back — in minutes. Nothing on the platform is configured by hand any more.

Since cut-over, the platform runs under our Managed IT Maintenance (SLA) contract. Prometheus and Grafana provide metrics and dashboards; alerts route to our 24/7 on-call rotation through Opsgenie and are classified P1–P4 with contractual response times. Every incident produces a written post-mortem, and the recovery procedure is re-tested on a fixed schedule — not assumed to work.