NOS Snapshot
We talked to: Joao Soares, Platform Engineering Lead, NOS
Industry: Telecommunications
Location: Portugal
Tech Stack: Hybrid (on-prem + Google GKE), Prometheus, Grafana, Kubernetes
NOS is one of Portugal’s leading telecom providers, serving millions with its network, internet, and entertainment services. Like many large enterprises, NOS transitioned to containerization years ago to drive agility and reduce infrastructure costs. But as Kubernetes usage grew, so did the complexity and difficulty of optimizing it. After years of trying to solve the problem manually, NOS turned to PerfectScale.
PerfectScale Results Snapshot
- 50% cost reduction on NOS’s largest cluster
- 0% idle resources using InfraFit and rightsizing automation
- Improved performance and eliminated SLA breaches
- Eliminated 2–3 full days/month of FinOps meetings
- Building cross-team trust in safe, intelligent automation
For NOS, cost optimization has always been at the heart of their infrastructure strategy. Almost a decade ago, that commitment led them to containerization. At the time, their systems relied on virtual machines, which were stable but inefficient and expensive. To reduce overhead and move faster, the team began experimenting with Docker and Rancher.
That journey eventually led to Kubernetes, which NOS rolled out across the company. Today, they operate a hybrid setup: telco-specific systems remain on-prem, while scalable workloads run on Google Cloud. The goal remains the same: deliver agility and performance while keeping infrastructure spend under control. As the architecture evolved, the complexity of managing it grew.
The Challenge:
Optimizing Kubernetes across environments became a major pain point. Teams at NOS were using tools like Prometheus and Grafana, but those only provided raw metrics. They lacked the visibility and recommendations needed to make safe, confident decisions, especially when performance or SLAs were on the line.
Without clear recommendations, engineers had to guess. After several failed attempts to rightsize manually, teams pulled back.
Manual optimization had become a liability. NOS needed a better way to manage resources across environments with more accuracy and less risk – performance excellence was a priority.
“Developers tried to make changes based on what they thought made sense, and it backfired.” said Joao Soares, Platform Engineering Lead, NOS.
Key challenges:
- Manual, time-consuming optimization: Observability tools only gave raw metrics, but no actionable guidance—which led to guesstimations and eventually errors.
- Limited cost visibility: Without granular insight into resource usage and spend, NOS struggled to identify overprovisioning and pinpoint areas of waste. This made it difficult to optimize costs without risking performance.
- Fragmented cost data: FinOps teams spent days each month piecing together fragmented data to build reports. The lack of centralized, reliable insights made it hard to prepare for budget reviews or spot issues in advance.
- Broken trust in optimization: Earlier efforts to optimize resources manually led to crashes and SLA breaches. This made teams wary of making changes for fear of performance issues.
At KubeCon Europe 2024 in Paris, Soares was looking for one thing—a safer way to cut Kubernetes costs. PerfectScale stood out immediately.
“I believed in the product the first time I saw it. I still show it to everyone,” said Soares. “It was the only solution that combined smart automation with real cost savings, without putting performance at risk.”
The Solution: Smart Automated Optimization, Built for Trust
PerfectScale was first rolled out in development clusters. Automation ran quietly for four months—with costs going down, and without a single issue or complaint.
With that success, NOS began automating platform-level components like the ingress controller, cert manager, and observability stack. In production, SREs still apply recommendations manually, but trust is growing. PerfectScale provides the contextual evidence they need to feel confident that changes won’t cause performance issues.
Results:
- Major cost reduction: NOS reduced costs by over 50% on their largest and most critical cluster—the main API cluster. Before PerfectScale, the cluster was heavily overprovisioned to avoid risk. With precise recommendations and automated rightsizing, NOS safely scaled back resources while maintaining performance.
- Performance improvements: PerfectScale uncovered overlooked issues, like out-of-memory kills and CPU throttling, then provided clear, targeted recommendations to fix them. Teams were able to increase resources where needed, without over-provisioning or guessing.
- Minimizing node utilization with smarter scaling: InfraFit helped NOS identify the optimal node types, enabling Cluster Autoscaler to scale more efficiently and in sync with user traffic.
“We’re seeing the sine wave we wanted—scaling up and down perfectly with user traffic.” Paired with automated workload rightsizing, several main node pools now run with 0% idle resource,” said Soares.
- Faster FinOps reporting: PerfectScale pulls all cost data into one place, so teams no longer spend days stitching together data. They now produce accurate, comprehensive reports in a fraction of the time, reclaiming two to three days every month.
“Now I walk into meetings knowing everything’s fine,” said Soares.
Company-Wide Adoption
PerfectScale now powers day-to-day decisions across NOS. It replaced scattered tools and guesswork with a single platform that the NOS teams trust. What started with platform engineering is now used across the company to improve performance, cut waste, and move faster— always maintaining the service quality.
This shared visibility has improved collaboration across teams and made Kubernetes optimization part of day-to-day operations.
Why PerfectScale
“I was specifically searching for a tool to help us try and reduce spend and help specifically developers and SREs to identify what the applications were doing and what resources they effectively needed,” said Soares. “PerfectScale was the right mix of doing what I wanted, with risk and resiliency safeguards, and the pricing was right.
PerfectScale has helped NOS rebuild internal trust in optimization—proving that it can be safe, effective, and even invisible when done right.
Ready to optimize like NOS?
Book a demo to see how PerfectScale can cut your Kubernetes costs, improve performance, and make optimization simple.
.png)
.png)
.png)
