Cast AI Nears Unicorn Status with New Backing from SoftBank and G2

“It’s all about GPU, compute, and electricity,” said Yuri Frayman, co-founder and CEO of Cast AI. “Our play is to ensure that we create efficiency, to be able to promote more workloads across GPUs. That is what we are about.”

Cast AI, a startup focused on optimizing cloud and AI infrastructure through Kubernetes automation, has raised $108 million in a Series C round. The funding is being used for further research and development and expansion into core markets, including the United States and Asia. The company now holds a post-money valuation close to $900 million, according to sources cited by TechCrunch.

The round was co-led by G2 Venture Partners and SoftBank Vision Fund 2. New investor Aglaé Ventures, backed by LVMH chairman and CEO Bernard Arnault, also participated, along with previous backers Hedosophia, Cota Capital, Vintage Investment Partners, Creandum, and Uncorrelated Ventures. The Series C follows Cast AI’s $35 million Series B round in November 2023, which had placed its post-money valuation at $300 million, according to PitchBook. Before this round, the company had raised just over $86 million.

Cast AI is headquartered in Miami, Florida, but its development operations are concentrated in Eastern Europe. Frayman described Cast AI as “a European company,” with most of its engineering teams located in Lithuania, Poland, Romania, and Bulgaria.

The company was founded in 2019 by Frayman, Leon Kuperman, and Laurent Gil. The founding team previously worked together at Zenedge, a cloud cybersecurity startup that faced challenges in managing cloud infrastructure costs during its growth phase. Zenedge was later acquired by Oracle. Prior to that, Frayman and Gil co-founded Viewdle, a machine learning startup that built early applications using Nvidia GPUs to train classifiers for image recognition. That company was later acquired by Google.

Cast AI was created based on the founders’ direct experience with cloud inefficiencies. While “AI” has always been part of the company’s name, the initial focus was on enabling more efficient and automated usage of Kubernetes-based workloads in multi-cloud environments. Over time, the platform expanded to support AI workloads, with increased demand from organizations managing compute-heavy operations.

The company now supports over 2,100 customers, including Akamai, BMW, FICO, Hugging Face, NielsenIQ, and Swisscom. Its platform integrates with all major cloud providers and also supports hybrid and on-premise workloads. The software analyzes cloud capacity in real time and determines the most cost-effective way to distribute workloads across CPUs and GPUs.

Cast AI cites internal research showing that only 10% of CPU and 23% of memory resources are typically utilized in enterprise environments. The company’s platform aims to reduce such underutilization through automated cost optimization, workload rebalancing, and resource right-sizing.

Kubernetes remains at the center of the company’s architecture and go-to-market strategy. The platform provides cost monitoring, optimization recommendations, and automated infrastructure changes using autoscaling, Spot Instance automation, bin packing, and other real-time resource management features. It also includes security functionality that scans for misconfigurations and vulnerabilities, and evaluates clusters against industry standards such as CIS Benchmarks.

In 2023, the company introduced the concept of Application Performance Automation (APA), a framework intended to replace static observability dashboards with automated, real-time infrastructure actions. “Companies are no longer satisfied with static dashboards and manual tuning knobs,” Frayman said. “They want automation that works out of the box, improves performance continuously, and gets smarter over time.”

The company doubled its customer count between 2023 and 2024. During that period, it also opened offices in India and Singapore, and was named both an IDC Innovator and a leader in Cloud Cost Management on G2.

Cast AI has formed partnerships with several large organizations. Frayman confirmed that the company is working with Crusoe Energy and is embedded in its infrastructure stack. It is also partnering with SoftBank on improving efficiency in AI data centers and is involved in a project led by OpenAI and SoftBank to develop AI infrastructure in Japan. “We are partnering with Crusoe, where we’re inside their stack, and we are partnering with SoftBank to be able to facilitate the efficiency in their AI data centers,” Frayman said. “We are partnering with the entire ecosystem.”

Tim Yap, Investment Director at SoftBank Investment Advisers, said, “Cast AI is setting a new standard for cloud efficiency at a time when infrastructure costs and demands are surging. We are excited to support their mission as they automate and accelerate how AI workloads are deployed at scale.”

Carl Fritjofsson, General Partner at Creandum, added, “Right now in the world, everyone is talking about AI agents. Cast was an AI agent before we started talking about that technology, you know. They’ve just been building this type of automation for a long time.”

According to Cast AI’s 2025 Kubernetes Cost Benchmark Report, overprovisioning and cloud resource waste remain persistent across the industry. The company is using its latest funding to further develop APA capabilities and extend its automation tools across more enterprise environments. Frayman said, “We are committed to making APA ubiquitous in DevSecOps.”