Inflection AI and Intel partnership accelerates model deployment for critical enterprise workloads

Inflection AI Inc. and Intel Corp. have launched a partnership designed to help enterprises tackle critical workloads and model deployment with AI.

The collaboration, which was announced earlier this month, facilitated Inflection AI's launch of an enterprise-grade AI system powered by Intel processors to deliver conversational AI capabilities while offering control, customization and scalability for large scale deployments.

"We've talked to a lot of companies about what their needs are and number one. What we heard was that there's a lot of dissatisfaction with the way in which the current commercial models are available for using automation and AI together because you have to go to the model provider," said Ted Shelton (pictured, right), chief operating officer of Inflection AI. "We started looking around for a partner who was trusted by the enterprise, who had an AI accelerator that we could use with our models, who we would be able to walk into the enterprise together and say, 'We have an opportunity for you to run your AI, and instead of you bringing the data to the model provider, we'll bring the model to you.' That was the genesis of this partnership."

Shelton spoke with theCUBE Research's Dave Vellante at UiPath Forward 2024, during an exclusive broadcast on theCUBE, SiliconANGLE Media's livestreaming studio. He was joined by Motti Finkelstein (center), chief information officer of Intel, and Kyle Short (left), director of generative AI and automation at Intel, and they discussed model deployment and how the two companies were addressing enterprise needs in generative AI and how UiPath Inc. fits in. (* Disclosure below.)

Inflection AI's consumer application was previously run on GPUs from Nvidia Corp. With the release of Inflection 3.0, this will now be powered by Intel's Gaudi 3 AI accelerator with instances on-premises or in the cloud.

"We rewrote all the acceleration of our product to be able to work on Gaudi 2 and Gaudi 3 AI accelerators using PyTorch and using Habana, and that was a big initiative there," Shelton said. "We also expanded the agentic framework to be able to include our partner UiPath so that right from the conversational interface, employees can trigger from AI the UiPath automations."

For Intel, the partnership with Inflection AI provides an opportunity to focus its technology on deployment of the model. Gaudi 3 offers users the benefits of improved price performance and 128GB of high bandwidth memory capacity to optimize generative AI.

"In Inflection AI, we're testing the use cases, what are the models?" Finkelstein said. "We care about the model, we care about the data, and the data is incredibly important to the enterprise. We have to make sure privacy rules are adhered to; we have to make sure security rules are adhered to. We are focused in AI on very large and very potentially, I'm going to say, game-changing items."

Those game-changing items will rely on partnerships such as the recent pairing of Intel and Inflection AI to keep pace with innovation. Generative AI has changed the landscape for business automation and created new opportunities for established tech companies to showcase what innovation can bring.

"The technology is advancing very, very fast," Short explained. "We're taking things through proof-of-concept and then realizing we need to take a different path and restart the POC because there's new capabilities that were just released that changed the game. All that generative AI is doing is widening the aperture of what is automatable in our processes. But it is the speed with which we are breaking down barriers and opening new doors that I think we are about to see the enterprise landscape fundamentally change."

Here's the complete video interview, part of SiliconANGLE's and theCUBE Research's coverage of UiPath Forward 2024: