AWS and Kubernetes Cost Optimization for IDEMIA with IT Specialist Support

IDEMIA needed a detailed analysis of cloud resource usage and identification of areas where the infrastructure was generating unnecessary costs.

The key challenge was to detect underused EC2 resources and incorrectly defined requirements for selected applications running in Kubernetes clusters.

The project required combining cost, infrastructure, and operational perspectives in order to reduce expenses without negatively affecting environment stability or application performance.

Edge One Solutions supported IDEMIA by providing IT specialists responsible for analyzing costs and resource usage in cloud environments.

The scope of support included analyzing data from AWS Cost Explorer, Metrics Server, and Prometheus, identifying underused resources on EC2 instances, and assessing the configuration of applications deployed in Kubernetes clusters.

The team also worked on selecting appropriate EC2 instance types, improving application resource requirement definitions, and preparing a mechanism that allowed selected environments to be temporarily paused when they were not required.

The team analyzed resource usage across the systems using data from AWS Cost Explorer, Metrics Server, and Prometheus.

Based on this analysis, underused EC2 resources and incorrectly defined resource requirements for some applications running in Kubernetes were identified. This made it possible to select more appropriate instance types and adjust application configuration to actual demand.

An internal solution was also implemented to temporarily pause environments that did not need to run continuously. As a result, the cost of test accounts was reduced more than threefold.

In cloud environments, infrastructure costs can grow quickly when resources are not regularly analyzed and aligned with the actual workload of applications.

The IDEMIA project shows that effective cost optimization requires combining financial data, technical metrics, and knowledge of how applications use resources in Kubernetes clusters.

For organizations using AWS and Kubernetes, a similar approach can help reduce unnecessary spending, improve autoscaling efficiency, and increase control over the costs of test and production environments.