In the world of technology, where the demand for intensive computing is soaring, especially with advancements in AI, the costs associated with training and running these models can quickly become a significant hurdle. For those familiar with the resource demands of processes like cryptocurrency mining or complex blockchain computations, the challenge of managing compute resources efficiently is well understood. This is where AI optimization becomes critical. Enter Cast AI, a startup dedicated to easing this resource headache by automating the optimization of AI, Kubernetes, and other demanding workloads. What is Driving the Need for AI Optimization? The explosion of AI development and deployment has placed unprecedented strain on existing infrastructure. Companies are grappling with the sheer volume of data and processing power required, leading to escalating costs and resource allocation complexities. Beyond AI, other modern workloads like those managed by Kubernetes also contribute significantly to compute demands. A key challenge is the underutilization of expensive computing resources. Cast AI highlights research indicating surprisingly low average utilization rates for CPUs (around 10%) and memory (around 23%). This inefficiency extends to GPUs, which are particularly vital and costly for AI tasks. In a market facing processor shortages, getting the most out of available hardware is paramount. Cast AI’s Solution for Kubernetes Cost and Cloud Efficiency Cast AI offers a platform designed to analyze an organization’s cloud and on-premise infrastructure to find the optimal cost-performance ratio for distributing compute workloads. By automating this process, they help businesses reduce waste and improve efficiency. While AI is currently a major growth driver, Cast AI’s roots and a significant portion of its business remain firmly in optimizing Kubernetes cost. The platform integrates with all major cloud providers, allowing customers to manage diverse environments effectively. Yuri Frayman, Cast’s CEO, emphasizes their core mission: “It’s all about GPU, compute and electricity. Our play is to ensure that we create efficiency, to be able to promote more workloads across GPUs. That is what we are about.” Their technology serves a diverse customer base, including large enterprises like Akamai, BMW, FICO, HuggingFace, NielsenIQ, and Swisscom, demonstrating its applicability across various industries facing high compute demands. Achieving Better GPU Utilization GPUs are the backbone for much of today’s advanced AI training and inference. However, like other compute resources, they are often underutilized, leading to wasted investment. Cast AI’s platform directly addresses this by intelligently managing workloads to improve GPU utilization. This focus on efficient GPU use is strategic, aligning Cast AI with major players in the AI infrastructure space. Notably, their recent funding round includes investors like SoftBank Vision Fund 2, placing them in the same portfolio as OpenAI and AI infrastructure provider Crusoe Energy. Frayman highlights these connections: “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 datacenters.” This integration into critical AI infrastructure projects underscores their role in maximizing the value of high-demand hardware like GPUs. Cast AI Funding and Future Plans Cast AI has successfully raised a significant $108 million in Series C funding. This substantial round follows strong growth and strategic partnerships. Sources indicate the funding values the company at close to $900 million post-money, nearing unicorn status. This marks a significant jump from their previous $35 million round in November 2023, which valued them at $300 million. Prior to this latest infusion, the company had already raised over $86 million. The Series C round was co-led by G2 Venture Partners and SoftBank Vision Fund 2. Other participants include Aglaé Ventures, Hedosophia, Cota Capital, Vintage Investment Partners, Creandum, and Uncorrelated Ventures. The capital will be used to accelerate R&D efforts and expand their business operations in core markets like the U.S., as well as internationally. Who is Cast AI? A Look at Their History and Reach Founded in 2019 by Yuri Frayman, Leon Kuperman, and Laurent Gil, Cast AI draws on the founders’ deep experience in technology and tackling resource challenges. Frayman and Gil previously built Viewdle, an early machine learning startup that utilized Nvidia GPUs for image analysis, later acquired by Google. Their subsequent venture, Zenedge (a cloud-based cybersecurity startup acquired by Oracle), highlighted the difficulties in controlling cloud costs at scale – an experience that directly inspired Cast AI. While officially based in Miami, Florida, Cast AI has a strong European presence, with significant development teams in Lithuania, Poland, Romania, and Bulgaria. In just three years of operation, the company has amassed over 2,100 customers, a testament to the widespread need for their optimization solutions. The Road Ahead: What This Funding Means The $108 million funding round is a clear indicator of investor confidence in Cast AI’s approach to solving critical compute challenges. At a time when AI and complex workloads are driving up infrastructure costs and resource demands, their platform for AI optimization, Kubernetes cost reduction, and cloud efficiency is increasingly vital. By enabling better GPU utilization and overall infrastructure management, Cast AI is poised to play a key role in making advanced computing more accessible and sustainable for businesses worldwide. To learn more about the latest AI optimization trends, explore our article on key developments shaping AI features.
