To get to the benefits that AI offers, organizations have to address their technology infrastructure in ways that are much broader than historical approaches. Senior analyst Greg Macatee joins host Eric Hanselman to delve into what's required and what enterprises are identifying in the recent Voice of the Enterprise AI and Machine Learning study. Enterprises are struggling with raising the success levels of AI projects. Over 60% report moderate to severe challenges in achieving AI success. Bringing together the computational power and the right quality data in the right locations can be complicated in the hybrid environments that more are operating. It's not just a matter of being more selective with use cases, AI requires a set of organizational skills that have to be honed. Starting small and iterating can reduce risk while building competency.

Infrastructure has to shift in new ways, as well. Data management processes that can build the necessary data pipelines to feed AI applications bring together a broader set of tech disciplines. There are new wrinkles in AI infrastructure ecosystems, with new providers looking to address supply chain constraints, like the Neocloud or GPU as a Service (GPUaaS) providers. Even hyperscalers are looking to them to meet surging demand in a tight market. Those new options offer new choices, but enterprises need to match them with their AI goals.

More S&P Global Content:

• Navigating the AI infrastructure landscape
• The path from LLMs to agentic AI
• Next in Tech | Ep. 225: Security for MCP

For S&P Global Subscribers:

• AI infrastructure strategies evolve amid widespread data challenges – Highlights from VotE: AI & Machine Learning
• Generative AI Market Monitor & Forecast
• AI infrastructure: Trends, thoughts and a 2025 research agenda

Credits:

• Host/Author: Eric Hanselman
• Guest: Greg Macatee
• Producer/Editor: Adam Kovalsky
• Published With Assistance From: Sophie Carr, Feranmi Adeoshun, Kyra Smith