Key Results
- 70% cost reduction through cluster consolidation, rightsizing workloads, and applying AWS Savings Plans & Reserved Instances.
- Simplified operations by consolidating clusters, AWS accounts, and environments, reducing management overhead.
- Faster deployments with GitOps using ArgoCD and Infrastructure as Code via Terraform.
- Improved security & governance with AWS Control Tower, SSO, GuardDuty, Security Hub, Istio Service Mesh, and runtime security tooling.
- Enhanced observability through cloud-native monitoring, logging with OpenSearch, Prometheus, Grafana, and tracing via OpenTelemetry.
- Ongoing Site Reliability Engineering services, enabling the client to operate a large-scale environment with a small internal team.
About the Client
A well-established Japanese financial services company, specializing in online brokerage and trading solutions for FX, ETFs, and CFDs. The firm provides secure, reliable, and scalable technology infrastructure to support institutional and retail clients, enabling efficient digital trading and investment operations across multiple markets.
The Challenge
- Legacy infrastructure spread across multiple on-premise and private-cloud data centers.
- High operational costs due to fragmented environments and underutilized resources.
- Complex deployment and scaling processes with manual operations.
- Multiple applications with separate clusters and AWS accounts, creating inefficiencies.
- Security and governance risks due to decentralized account and cluster management.
.avif)
Our Solution
1. Cloud Migration & Kubernetes Adoption
To eliminate infrastructure fragmentation and improve scalability, the team transitioned the client from EC2-based deployments to a fully containerized architecture powered by Amazon EKS on Amazon Web Services.
Applications were gradually containerized and modernized, enabling:
- Cloud-native deployment patterns
- Standardized release processes
- Improved scalability and resilience
- Reduced operational complexity
This phased migration ensured business continuity while transforming the core platform into a scalable Kubernetes-based environment.
2. Cluster & Account Consolidation
A major cost driver was infrastructure sprawl, too many clusters and isolated AWS accounts with underutilized resources.
The team re-architected the environment by consolidating clusters, merging accounts where appropriate and separating dev, staging, and production with clear governance boundaries
This significantly reduced idle capacity, simplified operations, and strengthened governance controls.
3. Strategic Cost Optimization
Rather than applying surface-level savings tactics, the team implemented structural cost engineering across compute and data layers.
Key initiatives included:
- Cluster and workload autoscaling to match real-time demand
- Spot instances for non-production workloads (achieving up to 70% savings in those environments)
- AWS Savings Plans for EC2/EKS workloads
- Reserved Instances for RDS
- Rightsizing databases and search infrastructure
- Migration to Graviton-based instances for improved price-performance
This multi-layered approach delivered up to 70% total infrastructure cost reduction.
4. Infrastructure as Code & GitOps Automation
All environments deployed and managed using Terraform and ArgoCD, ensuring reproducibility, consistency, and reduced manual effort.
5. Security & Governance Enhancements
The team introduced centralized governance with AWS Control Tower and SSO-based account access management. Additional security enhancements included:
- Service Mesh implementation via Istio
- Threat detection with Amazon GuardDuty
- Security posture monitoring through AWS Security Hub
- Centralized secrets management with AWS Secrets Manager
Consolidation also reduced the attack surface and improved auditability.
6. Observability & Monitoring
To ensure operational transparency at scale, a modern observability stack was implemented, including Prometheus, Grafana, and OpenTelemetry. This provided real-time visibility across services and significantly improved incident response and troubleshooting.
Ongoing Cooperation
Our team continues to manage the client’s cloud environment, providing reliable, secure, and cost-efficient operations without requiring a large in-house team.
Impact of our Work
The original client’s setup combined on-premise servers, traditional data servers, private and public cloud resources, creating fragmentation, duplicated costs and operational complexity.
We consolidated infrastructure into a containerized AWS environment, significantly reducing costs and operational overhead. Through Kubernetes adoption, cluster consolidation, autoscaling, and strategic use of AWS Savings Plans and spot instances, the company achieved major efficiency gains and stronger governance. The result was a more secure, scalable, and cost-efficient platform supported by ongoing managed SRE services.
