Read summarized version with
About project
Working time:
2024 – 2026
Industry:
FinTech, Payments
The service:
AWS Well-Architected Framework Review, Managed DevOps, Amazon EKS Modernization
Overview
A Canadian payment platform needed a reliable way to keep its cloud infrastructure secure, current, and ready to scale. The company lets North American merchants accept payments through popular Chinese payment systems.
The platform runs continuous live payment and settlement workflows on Amazon EKS, so infrastructure stability is critical to the business. The company did not need a full-time, in-house DevOps hire. It needed a trusted technical partner who could support day-to-day operations, respond to incidents, answer technical questions, and bring the right DevOps expertise whenever it was needed.
The organization partnered with Dedicatted, starting with an AWS Well-Architected Framework Review. Over two years, this grew into a long-term Managed DevOps partnership. The goal was not only to fix immediate gaps, but to keep the environment evolving through proactive improvements, optimization, and modernization. The partnership culminated in a zero-downtime migration of all production EKS clusters to a supported Kubernetes version.
The Challenge
The customer’s payment platform operated on a containerized AWS environment that included:
- 3 production Amazon EKS clusters
- 66+ applications managed through ArgoCD
- ALB, NLB, and MQTT endpoints handling live traffic
- Observability built on Grafana, Prometheus, CloudWatch, and Zipkin
The platform processed live payment and settlement workflows around the clock, which created several constraints for the business.
No dedicated in-house DevOps capacity: The company needed ongoing infrastructure expertise, but a full-time hire did not match the actual scope and variability of the work.
Unsupported Kubernetes version: All three production clusters ran Kubernetes 1.29, which had moved to extended support. This increased both technology risk and control plane costs.
Continuous operational demands: Incidents, technical questions, and infrastructure risks required fast, experienced engineering support to keep daily operations stable.
Zero tolerance for downtime: Payment and settlement services had to stay available at all times. Any infrastructure change carried direct business risk.
Want to get your copy of case study?
By submitting this form, you agree with our Terms & Conditions and Privacy Policy.
Technology & Approach
Dedicatted delivered the engagement as a continuous Managed DevOps partnership rather than a fixed-scope project. Assessment, ongoing operations, and strategic modernization were combined into a single model.
Step 1: AWS Well-Architected Framework Review
The engagement began with an AWS Well-Architected Framework Review (WAFR). It assessed the environment across cost, performance, reliability, security, and operational excellence. Based on the findings, Dedicatted:
- Developed and prioritized recommendations
- Created an actionable implementation plan for future improvements
- Identified the need for continuous infrastructure attention as the platform evolved
This review became the foundation for a long-term DevOps partnership.
Step 2: Ongoing Managed DevOps support
Dedicatted then integrated as an extension of the customer’s internal team. The team provided day-to-day infrastructure support, technical guidance, and engineering expertise across operational needs and new initiatives. It also investigated and resolved incidents to keep daily operations stable.
Step 3: Proactive optimization and risk mitigation
Beyond reactive support, Dedicatted continuously improved the environment, covering:
- Infrastructure, security, and reliability risks, identified and addressed before they could affect production
- Cloud costs, resource usage, and overall performance
- Monitoring and operational visibility, for faster troubleshooting and more reliable operations
Step 4: Zero-downtime EKS modernization
The partnership built trust over time, and Dedicatted took on increasingly strategic initiatives. The most significant was migrating all three production EKS clusters off an unsupported Kubernetes version without interrupting live payment and settlement workflows.
Instead of upgrading clusters in place, Dedicatted used a parallel-build approach. The team:
- Built new clusters on Kubernetes 1.35 with Amazon Linux 2023 node groups alongside the existing environments, allowing full validation before cutover
- Wrote dedicated cutover and rollback runbooks and validated them against live cluster states
- Migrated 66+ ArgoCD-managed applications step by step, along with ingress configurations, network rules, and observability stacks, verifying each component before routing production traffic
- Switched load balancer traffic across ALB, NLB, and MQTT endpoints
- Executed phased, independent cutovers for each production cluster in close coordination with the customer’s team
Once all traffic was running on the new clusters, the legacy clusters were decommissioned..
Business impact
The partnership delivered a stable, modern, and cost-efficient infrastructure, with no need for a full-time in-house DevOps specialist.
- Zero operational disruption: All three production EKS clusters moved from Kubernetes 1.29 to 1.35 with zero downtime to payment services.
- Reduced technology risk: The platform now runs on a supported Kubernetes version with modern Amazon Linux 2023 node groups, and the legacy clusters have been fully decommissioned.
- ~$13K in annual savings: Leaving extended support cut the EKS control plane rate from $0.60/hr to $0.10/hr per cluster. Across three clusters, that saves about $1.1K per month.
- A sustainable DevOps partnership: The customer gained a long-term, flexible partner who handles ongoing infrastructure needs, drives continuous improvements, and mitigates risks before they affect the business.
- Stronger visibility and resilience: Continuous work on monitoring, security, and architecture made the platform easier to troubleshoot and more resilient as it grows.
Is your EKS running on extended support?
If your clusters are several versions behind, or you’re paying extended support rates to delay an upgrade you can’t afford to get wrong, you’re not alone. Dedicatted can start with a review of your environment and a clear upgrade path, then stay on as your DevOps team for whatever comes next. Talk to our experts today
