K1x reducing cloud costs and streamlining Kubernetes operations with PerfectScale by DoiT
Meet K1x
K1x is a high-growth New Jersey startup building AI-powered tax software. Originally part of an enterprise organization, K1x was spun out to bring its innovative tax platform to market at scale. As part of the transition, the team migrated several Kubernetes clusters to Azure AKS. While the clusters offered a strong starting point, none of the engineers who originally ran them joined K1x. That left Jason Berk, the company’s lead platform operator, with the challenge of refining the infrastructure to support a nimble, fast-moving engineering team.
The Challenge
When Jason joined K1x, it quickly became clear that its Kubernetes environment was overbuilt for its needs. He discovered that node pools were significantly overprovisioned, often running at two to three times the necessary scale. “We had nodes in our cluster running at a minimum scale that was two or three times what we actually needed,” says Jason. “They never auto-scaled up, but we were using less than 10% of the resources. It was a huge waste.”
In addition to unnecessary infrastructure costs, K1x faced workload placement and visibility issues. Applications were deployed without alignment to Kubernetes best practices, and while monitoring tools like Prometheus offered raw metrics, they failed to provide actionable guidance. “It was like finding a needle in a haystack,” Jason shared. Expensive compute-optimized nodes were being used unnecessarily, leading to thousands of dollars in monthly waste.
The Solution
Jason knew he needed a faster, more intelligent way to evaluate and tune the environment. “I told my CTO: it will take me hours figuring it all out, what to measure, how to measure it, and whether I’m even doing it right, or we can just implement PerfectScale by DoiT. It’ll solve the problem faster and cheaper,” he says.
PerfectScale by DoiT delivered instant visibility and clear recommendations. Instead of manually reviewing metrics and configuration files, Jason could focus on refining the architecture and enabling his team. “PerfectScale by DoiT gives you the answer: ‘Here’s the workload, here’s the right settings,’” he says. “If those don’t work, it keeps refining the recommendations until it’s right.”
Jason relied on PerfectScale by DoiT daily, reviewing node utilization or drilling into workload settings to ensure everything was tuned correctly. With optimization work automated, K1x could prioritize other responsibilities. The shift wasn’t just about time savings, it changed how the team approached infrastructure. “PerfectScale made me ask the right questions, like, ‘Why are we running this workload on this node pool?’” Jason explains. “Within a couple of months, I shut those node pools down completely and saved us thousands of dollars a month.”
The Results
Within weeks of implementing PerfectScale by DoiT, K1x began seeing measurable improvements. Jason identified idle node pools and shut them down, immediately cutting cloud costs by thousands of dollars per month. Over-allocated workloads were right-sized with data-backed recommendations, improving efficiency without compromising performance. PerfectScale by DoiT also uncovered deeper issues, including memory leaks and one application that had silently restarted 4,000 times a month.
For Jason, the benefits extended beyond optimization. Managing Kubernetes for a startup with limited resources requires more than visibility. It demands confidence in decisions. “Within a month, we shut down entire node pools. The money we saved, thousands of dollars a month, went right back into our budget,” he says.
