Startup ScaleOps Inc. today announced that it has raised $130 million in Series C funding at a valuation exceeding $800 million.

Insight Partners led the investment with participation from Lightspeed Venture, NFX, Glilot Capital Partners and Picture Capital. ScaleOps’ total outside funding now stands at more than $210 million.

Allocating too little infrastructure to a cloud environment can lead to outages. Provisioning too much, in turn, generates unnecessary costs. Finding the optimal balance is highly difficult because the amount of hardware that applications require to work reliably regularly changes based on user demand.

ScaleOps offers a cloud platform that can automatically determine how much infrastructure should be allocated to each workload. That removes the need for engineers to perform the task manually. According to the company, its software boosts application reliability while cutting costs by up to 80%.

Many of ScaleOps’ features focus on optimizing cloud-based Kubernetes clusters. 

When traffic to a workload increases, Kubernetes creates a copy of the workload and directs a portion of user requests to that replica. The bigger the traffic spike, the more replicas Kubernetes creates. ScaleOps’ platform can optimize the number of replicas in a cloud environment based on a workload’s historical usage data and traffic forecasts.

Many organizations extend their Kubernetes clusters with an open-source tool called Karpenter. It automates some of the manual work involved in provisioning cloud instances. Additionally, Karpenter enables users to define a so-called disruption budget. It’s a number that determines how much of the hardware allocated to a workload can be deprovisioned without causing time.

ScaleOps’ platform can automatically set disruption budgets for Kubernetes clusters. Additionally, it enables companies to define custom workload scaling parameters. For example, a retailer could instruct ScaleOps to automatically create additional Kubernetes replicas at the time of day when traffic to its website is usually highest.

The software also optimizes cloud environments in other ways. It identifies opportunities to consolidate instances spread out over multiple instances into a single instance. Additionally, ScaleOps finds workloads that can be moved to spot instances, virtual machines priced significantly lower than standard cloud infrastructure. Such instances are more affordable because they’re often shut down abruptly by the cloud provider, which means they’re not suitable for many workloads.

Some infrastructure inefficiencies are the result of programming mistakes. Applications written in Java work by allocating a small amount of memory to each piece of code before it runs. Usually, they deprovision the memory when a code snippet completes its designated task. Failing to deprovision memory can lead to excessive RAM usage, which drives up costs.

ScaleOps’ platform includes dashboards that highlight Java applications with inefficient memory usage. According to the company, developers can use software to troubleshoot RAM allocation issues and thereby lower infrastructure costs. 

“Static allocation and manual tuning simply can’t keep up with the speed and complexity of modern production environments,” said founder and Chief Executive Officer Yodar Shafrir (pictured). “We built ScaleOps to change that, creating a new category of autonomous infrastructure management so that AI and cloud applications can run at full potential.”

The company will use its new funding to triple its headcount by year’s end. Additionally, ScaleOps plans to release new infrastructure features focused on artificial intelligence environments.

Photo: ScaleOps