The crushing of traffic entering training and running AI quickly became a major cost and resource headache for the organization. Today, CAST AI, a startup building tool that facilitates and optimizes workloads for AI and other tasks, raises key funding in the background of strong growth and partnerships with key players in the space.
The company will raise $108 million Series C, not only using both R&D, but also expanding to expand its operations in core markets such as the US and elsewhere. A source familiar with the deal told TechCrunch that the round has a company that costs nearly $900 million, as I understand it.
“We’re looking forward to seeing you in the future,” said Yurifreyman, CEO and co-founder of the cast. “Our play is to ensure that we generate efficiency so that we can drive more workloads on the GPU. That’s what we’re trying to do.”
In the context, when a $35 million cast was cast in November 2023, it was valued at $300 million in money post money per pitchbook. Before this latest round, the startup had just raised $86 million.
CAST AI is based in Miami, Florida, but “it’s a big location in Europe,” Frayman describes most of its development in Lithuania, Poland, Romania and Bulgaria as “European companies.”
The business has attracted 2,100 customers over the past three years. Companies such as Akamai, BMW, FICO, Huggingface, Nielseniq, and Swisscom use their technology to analyze cloud and on-premises capacity and find the best cost-effective ratio to distribute computing workloads across them. Frayman says it will integrate with what all major cloud providers and customers use.
When companies face a lack of processors to train and execute AI models, better resource allocation is needed. CAST AI claims, citing its own research, that only 10% of CPUs and 23% of memory are used, and the same goes for GPU usage.
This Series C highlights what else the startup is working on, both in size and in participants, and what others are working for.
G2 Venture Partners and SoftBank Vision Fund 2 are co-leading the round. Aglaé Ventures (an investment company of LVMH Chairman and CEO Bernard Arnault), as well as former backers Hedosophia, Cota Capital, Vintage Investment Partners, Creandum and uncorrelated ventures.
In particular, Frayman pointed out that oversubscribed rounds stabilize the same portfolio as Openai and AI infrastructure provider Crusoe Energy.
Freyman said his company already counts many of these companies as partners and customers. “We’ve partnered with Crusoe, who is in the stack, and now we can partner with SoftBank to promote efficiency in our AI data centers,” he said. “We are partnering with the entire ecosystem,” he added.
CAST AI has been talking to AI recently and doing a lot, but that wasn’t where the company started. The Ukrainian-born Frayman, who founded the company with Leon Kuperman and Laurent Gil in 2019, began his finance career before taking part in software development.
In 2006, he and Gill built what Frayman described as one of Viewdles, one of the “early machine learning startups.” So they built some of the earliest applications that train classifiers for image search using NVIDIA GPUs. “That’s how far forward you go in terms of understanding the power of machine learning,” he said.
The company will eventually be acquired by Google.
The three founders later worked on Zenedge, a cloud-based cybersecurity startup that was the cast’s inspiration. (Zenedge was eventually acquired by Oracle.)
Cast AI’s first use case came from its resource struggle experience. The cast has always had an “AI” in its name and spirit, but it’s about its application, particularly using the cloud and making allocations more efficient in Kubernetes workloads.
Kubernetes applications are still at the heart of startups, Frayman said, from both a revenue and spirit perspective (when you visit that site, you also see prominent messaging there). However, it is a surge in activity around AI, with all the talk and growth flowing from both customers and investors.
“CAST AI is setting new standards for cloud efficiency at a time when infrastructure demand is surged,” said Tim Yap, investment director at SoftBank Investment Advisers in a statement.
“Now, everyone in the world is talking about AI agents,” said Carl Fritjofsson, general partner at Creandum. “Before we started talking about the technology, the cast was an AI agent. You know. They’ve been building this kind of automation for a long time.”
Source link
