FinOps by design: strategic steering of a Cloud migration

Support for the migration of 250 applications to Kubernetes (GKE). Financial management rules embedded at CI/CD design time to prevent budget drift.

01 Challenges and context

Context and strategic stakes

During its digital transformation, the organization migrated a critical portfolio of 250 applications to Google Cloud Platform using an initial "lift-and-shift" approach on Compute Engine. The IT department aimed to achieve a technological leap by refactoring these applications to a Cloud-Native architecture on Google Kubernetes Engine (GKE). The primary challenge was to drastically control the costs of this dual migration, manage multi-geographic workload placement, and embed "FinOps by Design" practices directly into engineers' code.

02 Work carried out

Approach and methodology

During the critical migration phase, we ensured continuous budget monitoring between Compute Engine environments and the new GKE clusters. Our experts provided strategic Business Cases to inform architectural decisions: arbitration on Cloud region selection, optimal instance sizing, and aggressive utilization of Spot/Preemptible nodes for all non-production environments. A strict "Tagging by Design" policy was enforced from day one to guarantee accurate cost allocation for microservices.

03 Outcomes achieved

Results and indicators

35%

Infrastructure cost reduction

20%

Additional post-migration savings

100%

CI/CD chains with FinOps controls

  • Massive profitability: 35% reduction in infrastructure costs by abandoning the initial "lift-and-shift" paradigm in favor of GKE.

  • Continuous optimization: An additional 20% in savings post-migration through right-sizing operations.

  • Automated governance: Sustainable implementation of FinOps controls across 100% of CI/CD chains.