Manifacturing
All Case Studies
13 min

FinOps for on-premises and hybrid IT

-22%
IT spend reduction
The customer cut total IT spend by 22% in 18 months
-20%
License costs
Optimized license management (SAM)
20%
Time freed up by automation
The IT team reallocated 20% of its time to other work
Context

In September 2023, the IT Manager of a major group, a global leader in high-performance specialty alloys for critical sectors such as aerospace and medical, faced one of the most critical tasks in IT governance: setting the technology budget for the following year.

In recent years, the company had gone through a phase of strong expansion through the acquisition of other manufacturing companies. This had steadily enlarged its technology ecosystem, increasing infrastructure complexity and operating costs.

The IT infrastructure was characterized by:

  • data centers spread across the various production sites;
  • heterogeneous software estates inherited from the acquisitions;
  • numerous technology vendors;
  • difficulty in correlating IT costs with actual business needs.

The owners wanted more transparency on IT spend, more predictable costs, and fewer operational inefficiencies. To meet this challenge, the company decided to adopt a FinOps approach.
 

Problems

The initial assessment phase revealed several technical and organizational issues. Here is a brief overview.

  • No visibility into IT costs, because there was no structured system to correlate IT costs with organizational units or the services used.
  • Suboptimal management of software licenses, as decentralized license management had led over the years to uncontrolled growth in software costs.
  • Overprovisioning of infrastructure resources. Over time, the local data centers at the various sites had been designed for maximum availability. However, this had led to oversized servers, low hardware utilization, and difficult capacity planning.
  • Highly fragmented IT processes, which resulted in inefficient handling of internal requests, poor visibility into how work was distributed across teams, and the resulting difficulty in monitoring productivity.

Four fundamental transformations

  1. FinOps beyond the cloud: an early application of financial optimization frameworks to a mostly on-premises infrastructure, dispelling the myth that FinOps is exclusive to the public cloud.
  2. Post-M&A rationalization: lasting resolution of the technical debt and complexity left by past acquisitions, unifying governance of software licenses and hardware resources.
  3. End-to-end observability: moving beyond siloed management with the introduction of advanced full-stack monitoring, which eliminated costly overprovisioning while ensuring maximum business continuity.
  4. Cultural change (change management): a shift from a purely reactive approach to a proactive one, in which process automation and DevOps practices turned the time saved into new capacity for innovation.
Solution

To address these issues, Miriade defined a FinOps program structured in five macro-phases.

1. FinOps assessment and IT cost mapping

The first step was an in-depth assessment of the entire IT ecosystem. Through a workshop with the IT Manager and the heads of the various technology areas, we built a complete map of IT costs, covering software licenses, hardware infrastructure, maintenance contracts, and IT staff operating costs.

2. Software asset management and license optimization

To manage the software estate, we implemented Flexera One, a software asset management platform. Analysis of usage data identified unused licenses, overlapping tools, and non-optimized software contracts. Through rationalization and negotiation with vendors, the customer eliminated redundant licenses, optimized maintenance contracts, and consolidated its key platforms.

3. On-premises infrastructure optimization

The hardware infrastructure was analyzed with advanced observability and monitoring tools. We introduced the Dynatrace platform, used for full-stack infrastructure and application monitoring, resource consumption analysis, and automatic anomaly detection. The data analysis revealed significant room for optimization. We then launched a project to consolidate physical servers and virtualize the infrastructure, and redesigned the cluster architecture. The new infrastructure model reduced overprovisioning while maintaining high levels of resilience.

Process automation

A significant part of the project focused on improving internal IT processes. We introduced orchestration and IT service management (ITSM) tools: Camunda for business workflow orchestration, Jira Service Management for the service desk, and Jira Software and Advanced Roadmaps for project management.

Introduction of DevOps practices

Introducing DevOps practices made it possible to automate application deployment, standardize development environments, and introduce cost allocation models based on resource usage.
 

Technologies
  • Software asset management: Flexera One.
  • Observability and monitoring: Dynatrace.
  • Container platform: Red Hat OpenShift, an enterprise Kubernetes platform.
  • Process orchestration: Camunda.
  • IT service management: Atlassian Jira Service Management.
  • Project management: Atlassian Jira Software, Atlassian Advanced Roadmaps.