Delivering AI at Scale Without Missing a Beat

As agencies move artificial intelligence projects from prototype to production, they often realize they don’t have a clear path forward. One concern is infrastructure. While their initial deployments might be limited, agencies need to ensure the systems can scale over time, both maximizing existing investments and adapting to new requirements. Data is another challenge. Where does inferencing happen when some datasets reside on-premises but others in the cloud?

And how does the agency exercise strong governance over this complex environment? The solution is not to lock down that environment, reducing risk by eliminating options. Quite the opposite. Agencies must take an enterprise approach that provides strong, central management capabilities that make it possible to adopt new models as they emerge, said Jason Langone, Senior Director of Global AI Business Development at Nutanix.

“It seems like there’s a new model every other day that’s better than the one that was there,” he said. “So, how do you ensure you’re not locked into a particular hardware accelerator or model provider? How do you have freedom of choice?”

In this video interview, Langone discusses how agencies can deploy AI at the enterprise level without compromising reliability, flexibility or security. Topics include:

• Developing a hybrid accelerator strategy for AI infrastructure
• Deploying AI models where the data resides, whether in the data center or at the edge
• Giving IT staff the tools they need to deploy and manage AI initiatives effectively and efficiently