In April, Snowflake Inc. announced that it would bring Meta Platforms Inc.’s Llama 4 large language models to Snowflake’s Cortex AI. The models, trained with large amounts of unlabeled text, image and video data, were designed to accommodate a wide range of use cases and developer needs.

Meta’s Amit Sangani and Snowflake’s Dwarak Rajagopal talk with theCUBE about large language models.

The partnership with Snowflake was part of a larger effort on the part of Meta to make its models available in the open-source community.

“The whole purpose of open sourcing these models was to make sure we democratized a developer sitting anywhere in the world to be able to get access to the same infrastructure,” said Amit Sangani (pictured, left), senior director of AI platform engineering at Meta. “They can download these models, they can fine tune the models. That’s the power. You’re basically elevating the innovation across the world.”

Sangani spoke with theCUBE’s Rebecca Knight and George Gilbert at Snowflake Summit, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. He was joined by Dwarak Rajagopal (right), vice president of AI engineering and research at Snowflake, and they discussed the evolving role of models in enterprise AI. (* Disclosure below.)

Business impact through large language models

For Snowflake customers, the benefits from access to Llama models are part of the data warehousing company’s interest in continuing to broaden the scope of capabilities available through Cortex AI. The company made a number of additional enhancements to Cortex this week during its annual conference, and Rajagopal described how Snowflake customers have been leveraging the platform.

“Tripadvisor [has] been using Llama models via Cortex to actually power personalized recommendations, travel recommendations for their users,” Rajagopal noted. “They have seen positive customer engagement but also big business impact as well.”

The collaboration between Snowflake and Meta also highlights the increasingly critical role that models are playing in agentic AI. As enterprises adopt intelligent agents to execute specific business tasks, it becomes more critical that the underlying infrastructure can support an increasingly complex process.

“One of the biggest innovations happening in the industry is around agent existence, where a complex workflow is broken down into multiple steps, and the model is the foundation of how you do that,” Sangani explained. “By next year, if we are able to reduce the hallucinations significantly, and models can do a multi-step process successfully, I think that’s a big win for the industry because then that opens up tremendous opportunities for AI.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Snowflake Summit:

(* Disclosure: TheCUBE is a paid media partner for Snowflake Summit. Neither Snowflake Inc., the primary sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE