20% cost reduction
in Costs
Improved user experience
with lower latency.
20 employees
trained in FinOps
Improved Budget
of AWS EKS costs.
Ready to start?
Challenge
At the outset, the organization was running several AWS EKS clusters in a highly dynamic environment… but cloud costs were growing faster than the business activity. The teams already had a few optimization levers in place (auto-scaling, instance sizing), but without a common framework or consolidated visibility, optimization efforts remained occasional and difficult to scale.
The first challenge was therefore to regain control over cloud consumption: fine-tune auto-scaling parameters and right-size EC2 instances to eliminate overprovisioning without impacting performance. At the same time, it was necessary to create a single source of truth: a data warehouse capable of centralizing all expenses related to the various EKS clusters.
Once reliable data was in place, the objective became clear: turn costs into decisions. We needed to calculate and consolidate expenses by scope, then define cloud financial KPIs by business unit to monitor, prioritize, and measure the impact of optimization initiatives.
Finally, to ensure the initiative would be sustainable over time, the challenge was not only technical: it was essential to bring the teams on board. More than 20 employees were trained in FinOps practices, from design through to production deployment, to establish long-term governance and prevent costs from drifting again.
Solution
To move from “ad hoc optimization” to a sustainable approach, we first established a clear framework: a target FinOps process, measurable objectives, and governance that was simple to maintain over time. A Leonys consultant, with expertise in FinOps and AWS, was appointed to lead the implementation and oversee day-to-day operations.
Very quickly, the challenge was to embed FinOps where decisions are actually made: within projects. We therefore integrated the FinOps approach into the entire lifecycle, from the architecture phase through to post-production support. Concretely, each architecture document was enhanced with a FinOps deliverable: an estimate of implementation costs and recurring costs, enabling spending to be anticipated before it became an issue.
Next, to ground the approach in reality, we started with a proof of concept involving a pilot team. Using billing data (via AWS Cost Explorer), we built an initial clear view of spending, identified the highest-priority quick wins, and quickly demonstrated the value of the approach — without adding unnecessary process overhead.
Finally, the key to success was knowledge transfer: rather than keeping ownership of FinOps with the expert, we progressively empowered teams, equipped them with the right rituals and tools, and trained employees in FinOps best practices. The result: an organization capable of making informed decisions, measuring outcomes, and continuously optimizing costs, with aligned trade-offs across technology, product, and finance teams.
