RelationalAI Becomes the First AI Coprocessor that Runs Securely within Snowflake

RelationalAI, the industry’s first AI coprocessor for cloud platforms and language models, is announcing that its cloud native, AI coprocessor—compatible with the Snowflake Data Cloud—is available for preview at Snowflake’s annual user conference, Snowflake Summit 2023. Furthermore, Relational AI is being named a launch partner for Snowflake’s new Snowpark Container Services, designed to expand the focus of Snowpark to aid organizations running third-party software.

RelationalAI tackles the challenge of having to replicate data and use technologically specialized tools—such as navigational graph databases, rule engines, and solvers—to build intelligent solutions. RelationalAI, with Snowflake, enables developers to create solutions where the data lives, rendering rampant complexities and costs nonexistent, according to the company.

This technology further addresses a gap in the market, where solutions did not exist as both relational and cloud native, making it difficult to achieve the benefits that RelationalAI offers.

“You have too many components you have to bring together when you build intelligence into an application,” said Molham Aref, CEO of RelationalAI. “And each part has its own architecture and its own paradigm; they're not always data-centric, and most of the effort goes in writing glue between all of these parts…and so, in a world that's moving to be more data-centric…we thought the right place to add all that capability is into the database.”

As the first AI coprocessor of its kind, RelationalAI separates storage and compute, offering zero-copy cloning and workload isolation, built on a trusted, proven relational paradigm. The solution extends Snowflake by empowering customers to innovate  their data in one place—within their secure, governed Snowflake account. This results in increased performance, streamlined management, and creates new opportunities for developmental innovation based on the closeness of their data, according to the vendors.

“The relational paradigm is proven, it's trusted; most data in most enterprises is stored in systems like SQL, or SQL-based systems like Snowflake. And we're here to say that we can make it work for other use cases,” said Aref. “That way, you don't have to go to something more complicated to get the intelligence built into your infrastructure and your applications.”

In Snowflake’s effort to create a broader reach for Snowpark, RelationalAI benefits from streamlined and simplistic operation within Snowflake accounts. This dramatically increases the value of an enterprise’s Snowflake data, without the need to comprise its security.

Many might “think this is too good to be true, because they're so conditioned on the fact that you need different database technologies for different types of workloads,” explained Aref. “Just this idea of, ‘hey, it can all be available to you inside of your favorite environment,’ is going to get a very interesting and positive reaction…because the alternatives are just so bad.”

Despite being available for preview, RelationalAI has been in production with some early, large customers for nearly three years—and the reactions emphasize its value-driving capabilities.

“They [the early customers] appreciate using our system to develop very, very sophisticated applications, accounting, and so on. This [technology] is already validated in the context of a small number of large accounts and a small number of small accounts,” concluded Aref.

A demo of RelationalAI’s capabilities will take place at booth #2132-C at Snowflake Summit 2023, this week, in Las Vegas. Interested parties can also sign up for a preview of RelationalAI on the Snowflake data Cloud by contacting the company here.

To learn more about RelationalAI, please visit