Snowflake Unveils Cortex: A Unified AI Tool for Analysts and Developers
Snowflake Cortex: Empowering Data Management with AI Integration
In the realm of cloud-based data storage, Snowflake stands as a go-to platform for companies managing their data. With the burgeoning interest in generative AI and expansive language models, users are seeking seamless ways to integrate these technologies. Enter Snowflake Cortex, a fully managed service that caters to the needs of both business analysts and developers, streamlining the incorporation of AI-driven applications on the Snowflake platform.
Tailored AI Tools for Business Analysts
Tailored to different roles, Snowflake Cortex offers business analysts access to a suite of AI tools, leveraging Snowflake’s proprietary Large Language Models (LLMs). This empowers analysts to effortlessly interact with data stored in Snowflake, enhancing efficiency and speed. Meanwhile, developers can utilize Cortex to construct generative AI applications atop Snowflake data, tapping into capabilities introduced through the Streamlit acquisition last year.
Sridhar Ramaswamy, SVP of AI at Snowflake, highlighted the core component of this advancement, Snowflake Cortex, stating, “At its core, we are bringing advanced search, as well as large language models right into the heart of Snowflake.” The objective is to deeply integrate these advanced features within Snowflake, enhancing productivity for power users and analysts entrenched in the platform.
For business analysts, Snowflake Cortex introduces Document AI, facilitating the extraction of data from unstructured documents such as PDFs and analyst reports. This feature simplifies the process for analysts, enabling them to extract structured values without requiring specialized programming or language model knowledge. Universal search, a capability acquired through the Neeva acquisition, allows users to search across all Snowflake data and the marketplace efficiently.
The third pivotal feature is Snowflake Copilot, designed to convert plain language questions about Snowflake-stored data into SQL queries. This potentially saves analysts significant time by bypassing the need to familiarize themselves with data structures to formulate meaningful queries.
Developer Empowerment with Snowflake Cortex
Developers, on the other hand, can leverage Snowflake Cortex to swiftly build applications using pre-existing models or opt for a more customized approach. Access to external Large Language Models, such as open-source offerings or those from cloud partners like Amazon Bedrock and Azure OpenAI, provides additional flexibility. The deployment efficiency of applications is further enhanced through the use of Snowflake Container Services, announced in June, which allows for containerized workloads.
Snowflake Cortex is a strategic move to unlock the full potential of data stored in Snowflake, whether it’s through searching, querying, or application development. As of now, Cortex and its core features are in a private preview phase, with no specified timeline for wider availability.