Client: Regional Telecom Provider (Asia)
Tools: EKS, Helm, Karpenter, Prometheus, Grafana, GitLab CI
Challenge:
Kubernetes environments were provisioned manually with no autoscaling or monitoring standardization. Node underutilization caused rising cloud bills.

Solution:

• Automated cluster creation using IaC (Terraform + EKS blueprints)
• Used Karpenter for dynamic autoscaling
• Created Helm charts for all microservices with GitLab CI-based release pipelines
• Standardized logging and monitoring with Prometheus + Loki + Grafana
• Deployed cost dashboards and alerting pipelines

Outcome:
Reduced ops effort by 70%, cut infra cost by 25%, and improved observability across 50+ workloads.