Companies use Snowflake to store your data in the cloud. With growing interest in generative ai and large language models, customers are looking for ways to get up and running with the technology quickly. Today, the company announced Snowflake Cortex, a fully managed service designed to help both business users and developers work with ai-powered applications on the Snowflake platform.
It has a couple of purposes, depending on its function. For business analysts, it provides access to several ai tools built on Snowflake’s custom LLMs to make it easier and faster to interact with data stored in Snowflake. For developers, it helps them build generative ai applications on top of data stored in Snowflake, in part by leveraging a capability that came to Snowflake with its acquisition of Streamlit last year.
“In essence, we’re bringing advanced search as well as large language models right to the heart of Snowflake with a new component we’re calling Snowflake Cortex,” said Sridhar Ramaswamy, senior vice president of ai at Snowflake, in a press roundtable last week. pass.
“We want to create these advanced features, which are increasingly a requirement for the modern enterprise, and integrate them deeply into Snowflake, so that our power users, the analysts who spend almost all their time in Snowflake, become so much more. more productive,” said Ramaswamy, who came to the company as part of the Neeva acquisition earlier this year.
While developers can leverage Cortex to create generative ai applications, the company provides several advanced elements out of the box to help analysts take advantage of generative ai. The first is Document ai, a way to extract data from unstructured documents like PDF files and analyst reports and query that information. “What Document ai does is make it easier for an analyst without any specialized knowledge of programming or large language models to extract these structured values from these documents and put them into a table,” he said.
In practice, this allows analysts to ask questions about the unstructured data stored in these documents.
The second feature they’re adding is universal search, the capability that came to Snowflake when it acquired Neeva in May. “Search, as many people realize, is the foundation for doing interesting things with language models, and we’re exposing the core of search on Snowflake objects,” he said. This allows users to search all of their Snowflake data and the Snowflake Marketplace to locate data or apps they’ve created.
The third key piece of Cortex analytics is Snowflake Copilot, which takes plain language questions about data stored in Snowflake and turns them into SQL queries. If done correctly, this could save analysts a lot of time spent getting familiar with the data and column structure to create meaningful queries.
For developers, they can quickly build apps using Snowflake models, or for those who want more control over the entire process, they can build more custom apps with access to external LLMs like open source offerings, or those from cloud partners like Amazon Bedrock and Azure OpenAI. They can also take advantage of Snowflake Container Services announced in June to deploy applications more efficiently as containerized workloads.
Snowflake Cortex is part of a larger plan to make data stored in Snowflake work in different ways, whether searching, querying, or building applications. For now, Cortex and its core features are in private preview. The company has not indicated when it will be more widely available at this time.