The challenge
A bank’s payment platform handled more than $10 million in transactions every day - the kind of system where downtime is measured in lost trust, not just lost minutes. As volume grew, latency crept up and the infrastructure needed to be hardened for both speed and resilience.
Our approach
- Re-engineered the Kubernetes infrastructure underpinning the payment services for stability under heavy, spiky load.
- Introduced a service mesh across 40+ microservices to standardise and speed up how services talk to each other.
- Tightened cluster lifecycle management and disaster-recovery procedures so failures could be handled without customer impact.
- Mentored the engineering teams on cloud-native operating practices to keep the gains in place long-term.
The results
Inter-service latency reduced by around 25%.
Zero unplanned downtime across the period of work.
A more resilient platform able to absorb growth in transaction volume.
Engineering teams upskilled on running mission-critical Kubernetes workloads.
Technologies
- Kubernetes
- Istio service mesh
- AWS
- Microservices
- Disaster recovery
Client kept confidential. Delivered in a senior architecture / engineering role.
Back to all work