The quickly evolving landscape of artificial intelligence and data management is witnessing game-changing advancements, thanks to the collaboration between leading tech giants.

As companies such as Snowflake Inc. and Nvidia Corp. increasingly rely on AI to drive insights and efficiency, integrating sophisticated AI models into secure data platforms has become crucial. This synergy, in the form of Nvidia NeMo, ensures that enterprises can deploy AI in stable, production-ready environments, unlocking new potential across various sectors.

“Our partnership started last year,” said Kari Ann Briski (pictured, right), vice president of generative AI software and product management at Nvidia. “We announced the release of our [Neural Modules], which is for the customization of generative AI models into the Snowpark container services. This year, we’ve taken it even further. We’ve deepened our partnership. What we’ve done is that we’ve integrated Nvidia NeMo Retriever as a [Nvidia Inference Microservices] into Cortex AI.”

Briski, along with Vivek Raghunathan (left), engineering at Snowflak,e spoke with theCUBE’s Dave Vellante and Rebecca Knight at Data Cloud Summit, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the integration of advanced AI models into strong data platforms and how that is becoming essential for businesses looking to leverage AI effectively. (* Disclosure below.)

Strengthening partnerships for enhanced AI

Nvidia and Snowflake have significantly deepened their collaboration to advance the capabilities of enterprise AI. This integration aims to streamline the deployment of AI applications by embedding them directly within enterprise data environments, which play a critical role in the efficiency and security of these advancements, according to Raghunathan.

“Cortex is the idea that you have an offering that makes AI really easy for our customers to use,” he said. “They can run it in production; it is safe and it is secure within their governance boundary.”

It is important to ensure that AI-driven insights are both immediate and actionable, Raghunathan explained. This approach is reflected in the development of Cortex AI, which provides tools such as Cortex Inference, Cortex Search, Cortex Analyst and Cortex Fine-Tuning, catering to diverse enterprise needs.

“Nvidia truly understands that a great AI strategy is powered by a great data strategy,” he added. “To really have these AI applications run in production, you need to run them next to the data.”

Democratizing AI for enterprise use through Nvidia NeMo

One of the core goals of the Nvidia-Snowflake partnership is to democratize AI, making it accessible and beneficial across various levels of enterprise operations. This democratization is evident in the practical applications of AI within enterprises. For instance, the integration of the NeMo Retriever NIM facilitates the creation of interactive chatbots and sophisticated data analysis tools.

“You ask a question; it goes and retrieves some information out of your vector database similarity search. They bring it back and then you send it to the generative model,” Briski explained. “The accuracy of that retrieval process really matters.”

Efficiency is also a big part of in AI deployment. This focus on efficiency is crucial for reducing the total cost of ownership and maximizing return on investment for enterprises adopting AI solutions, according to Briski.

“Our goal is to use less of the chip because we want to give you the most efficient inference possible,” she said. “We’re always trying to optimize it to get down to the most optimal model. The total cost of ownership goes down and you get more return on investment.”

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

(* Disclosure: Snowflake Inc. and Nvidia Corp. sponsored this segment of theCUBE. Neither Snowflake, Nvidia nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE