About Maxar Technologies
Maxar Technologies is a leading aerospace innovator and a critical provider of geospatial intelligence to the U.S. government, powering over 90% of its geospatial data. As a trusted partner of NASA and a key provider to defense and intelligence agencies, Maxar operates in high-stakes, mission-critical, air-gapped environments where uptime, performance, and reliability are essential.
The Challenge
Evolving Needs in a Complex Kubernetes Environment
Maxar runs dozens of Kubernetes clusters, including air-gapped environments running critical production workloads. As their infrastructure scaled, so did the complexity of managing compute resources efficiently across varied environments.
The Platform team identified an opportunity: while developers were setting resource requests to ensure performance headroom, many workloads were consistently over-provisioned. These conservative configurations, often inherited from legacy systems like Cloud Foundry, didn’t always reflect real-time usage patterns in Kubernetes. Limited visibility and the manual effort required for tuning made optimization difficult to prioritize.
Additionally, operating across secure, air-gapped environments meant Maxar needed a platform that could deliver application context-aware automation, without relying on outbound connectivity or complex integrations.
“Before ScaleOps, we had almost no control over what developers were deploying into our production clusters. They were just overprovisioning everything, requesting four CPUs for workloads that barely used half a core. We tried building dashboards so teams could track actual usage, but we’d still have to chase them down constantly to tune things. And even then, nothing changed.” – Jeff Burger, Lead Platform Engineer
From Insights to Real-Time Automation
Maxar didn’t need another observability tool or recommendation engine. They needed a platform that could act automatically and in real-time.
Manual recommendations weren’t enough. What they wanted was real-time, context-aware automation that could safely and continuously rightsize workloads in production, with zero developer effort and no risk to critical services.
“We came in looking to save costs, but not at the expense of performance. With ScaleOps’ automated resource optimization, we got both, and saw dramatically high cost savings without compromising on reliability.– Jeff Burger, Lead Platform Engineer
Partnering with ScaleOps
Effortless Optimization at Scale
By deploying ScaleOps, Maxar enabled real-time, context-aware optimization across all Kubernetes environments, including air-gapped systems. The platform continuously adjusts CPU and memory requests based on actual usage, live cluster conditions, and workload context.
With no dashboards to monitor or tickets to process, the Platform team achieved optimization without disrupting developer workflows. Developers could continue focusing on innovation, while ScaleOps handled resource efficiency in the background.
With ScaleOps, that burden disappeared.
“ScaleOps eliminated the friction between platform and app teams. It’s automated, it’s trusted, and it works in production, without the usual overhead.” – Chris Ernst, Sr. Staff Automation Engineer
The Results
100% Automation in Production. 62% Less CPUs requested. 40% Less Memory Requested. Zero Disruption.
By partnering with ScaleOps, Maxar reduced CPU requests by 62% across all production environments. These savings translated directly to AWS cost reductions, without a single manual change.
From day one, ScaleOps delivered value with zero-touch onboarding, no manual tuning, no dashboard monitoring, and no customization needed. Within minutes, workloads were automatically rightsized in production.
Multi-cluster automation capabilities enabled Maxar to manage optimization across highly distributed environments, including air-gapped systems, with consistency and confidence. Now, every production workload is automatically rightsized 24/7, while also improving performance and reliability.
By partnering with ScaleOps, we significantly reduced our cloud spend without touching a single workload manually. In production environments as complex as ours, that kind of impact is rare. It’s like having a full-time engineer whose only job is to make sure every workload has the resources it needs, 24/7. – Jeff Burger, Lead Platform Engineer
