The artificial intelligence wave continues to surge, but most companies are quietly hemorrhaging money behind the glossy headlines. In server rooms across the globe, top-dollar GPUs wait in silence; over-provisioned systems idle at half-capacity. The bill for all this wasted potential arrives, as usual, with the cloud provider’s logo. ScaleOps sees it differently—misuse, not scarcity, drives this profligacy.
This week, ScaleOps—a startup laser-focused on taming the chaos of enterprise compute—announced it had brought in $130 million in fresh funding, rocketing its valuation to a cool $800 million. The round, its Series C, was steered by Insight Partners and drew renewed backing from its veteran supporters: Lightspeed Venture Partners, NFX, Glilot Capital Partners, and Picture Capital. Their bet? ScaleOps’ promise to slash the monstrous infrastructure bills that haunt cloud and AI leaders—by as much as 80%.
The story of ScaleOps begins in 2022, when Yodar Shafrir—a hands-on technologist with a stint at Run:ai, the orchestration startup later snapped up by Nvidia—grew tired of watching sophisticated workloads choke under their own weight. Tools like Kubernetes, now standard in the field, offer brute power, but remain surprisingly rigid. You set up configurations, then cross your fingers as demand veers, often unpredictably, sending valuable resources into limbo.
Shafrir puts it simply. “Working with DevOps teams at Run:ai, I saw the same headache repeat itself—no matter how advanced the tools, wrangling production workloads became an endless struggle, especially as AI inference took center stage,” he told TechCrunch. “But soon it became clear: this wasn’t only a GPU problem. The mess extended to compute, memory, storage, networking—the whole ecosystem teetered on the same fragile ground. Teams, again and again, failed to deploy resources wisely.”
For these DevOps engineers, every outage sets off a bureaucratic relay race—frantic emails, status calls, and ultimately, little progress. Visibility tools paint the shape of the problem, but rarely reach for solutions. That gap, ScaleOps realized, was a cavernous commercial opportunity.
ScaleOps stepped in with an answer: software that senses what every application needs in the moment and adjusts infrastructure in real time—no manual tuning, no late-night firefighting. “Kubernetes is an incredible foundation,” Shafrir concedes. “But it’s both a blessing and a curse—the flexibility demands humans to constantly reconfigure. That doesn’t fit a world of ever-shifting, high-stakes workloads. You need something attentive, intelligent, truly contextual.”
Unlike rivals such as Cast AI, Kubecost, or Spot, which can stumble without full awareness of changing production needs, ScaleOps claims their approach is fully autonomous from day one—context-aware and ready to run, zero hand-holding required. The aim is trust: let teams step out of the loop, and infrastructure adapts behind the curtain.
ScaleOps, headquartered in New York, now serves enterprise giants worldwide—companies whose operations ride on Kubernetes and stretch from Silicon Valley boardrooms to datacenters across Europe and the subcontinent. Among their customers: Adobe, Wiz, DocuSign, Salesforce, Coupa—names synonymous with scale, where a single hour of downtime is an expensive problem.
Their growth figures speak loudly: over the past year, revenue surged 450%, and staff numbers have tripled—with another hiring spree about to unfold. The Series C infusion comes only 18 months after a $58 million Series B. Their war chest now stands at roughly $210 million.
As the hunger for AI continues to spiral, ScaleOps plans to broaden its product suite, relentlessly pursuing the holy grail of fully autonomous, self-healing infrastructure. It’s a race to turn the whispers of efficiency into hard, measurable results before costs spiral completely out of control.
In a cloud world flooded with buzzwords, ScaleOps is staking its future on substance—turning relentless mismanagement into precision, and expensive waste into a competitive edge.
