Most companies do not achieve ROI from generative AI due to inadequate infrastructure and poor integration into operations.

Although companies in the US have invested between 30 and 40 billion dollars in generative AI, 95% of them have not reported a measurable ROI (Return on Investment). A MIT study highlights that the issues are not related to human talent, but to technological limitations, such as the inability to learn continuously and to integrate into essential workflows.

Most AI projects fail before reaching production due to poor integration into the business environment and uncontrolled use of tools by employees.

Björn Kolbeck, CEO of Quobyte, emphasizes that storage is an essential component often overlooked, and traditional storage systems are not designed to support large workloads. The proposed solution is a scale-out architecture, which allows for horizontal expansion of storage to support AI growth. Kolbeck warns that without adequate storage infrastructure, the potential of AI will never be realized.