Harnessing AI with Snowflake's Cortex LLM Functions
Written on
Introduction to Snowflake's Cortex LLM Functions
Snowflake has introduced its Cortex LLM Functions, aimed at enriching AI capabilities within its platform. As competition intensifies among data warehousing services, it’s no surprise that Snowflake has rolled out LLM functionalities to keep pace with rivals like Google and Microsoft, who have also launched similar offerings.
In the latest release from March, Snowflake announced the preview of these Cortex LLM functions, which provide immediate access to an array of features powered by cutting-edge large language models. These models are fully hosted and managed by Snowflake, eliminating the need for user setup while operating under Snowflake's governance and security protocols. The available functions include:
- COMPLETE: This function generates a response that completes a given prompt, accommodating both single prompts and dialogues consisting of multiple exchanges.
- EXTRACT_ANSWER: This function retrieves answers from unstructured data when a question is posed, provided the answer is present in the dataset.
- SENTIMENT: It assesses the sentiment of a text, returning a score from -1 to 1, indicating positive or negative sentiment.
- SUMMARIZE: This function condenses the provided text into a summary.
- TRANSLATE: It translates text between supported languages seamlessly.
As mentioned, Snowflake Cortex is integrated into the Snowflake ecosystem, ensuring ease of use for customers. Furthermore, Snowflake is unveiling novel LLM-powered experiences that leverage Cortex as the foundational service.
New LLM-Powered Experiences
The first video titled "How to start using Snowflake Cortex LLM functions: Python" provides a comprehensive overview of how to implement these functions effectively.
These holistic experiences include user interfaces and high-performance LLMs that are entirely managed by Snowflake Cortex, making them optimal for business teams and analysts. According to Snowflake, these enhancements are designed to boost productivity across the Data Cloud. The new LLM experiences introduced by Snowflake encompass:
- Snowflake Copilot: Utilizing advanced language models, Snowflake Copilot acts as a user-friendly assistant for routine coding tasks. By employing natural language, users can easily query their data, craft SQL queries tailored to specific datasets, adjust queries, and filter insights.
- Universal Search: Snowflake has developed a state-of-the-art search function powered by LLM technology, allowing users to quickly find and utilize the most relevant data and applications for their needs. This extensive search capability covers a user's entire Snowflake account, including databases, views, and soon-to-be-released Iceberg Tables. Additionally, users can access data and Snowflake Native Apps on the Snowflake Marketplace.
- Document AI: As Snowflake's first venture into LLM experiences, Document AI enables businesses to leverage language models for efficient content extraction from documents, such as invoice amounts or contract terms. Users benefit from a visual interface and natural language interactions to refine results, facilitating informed business decisions while enhancing operational efficiency in a secure and scalable environment.
The second video, "Live: Snowflake Cortex ML Functions," discusses the practical applications and benefits of these new features in real-time scenarios, showcasing their potential impact on data management.
Conclusion
Snowflake's Cortex LLM Functions represent a significant advancement in data warehousing capabilities, providing users with powerful tools to enhance productivity and streamline data management. With these innovative features, Snowflake continues to position itself as a leader in the competitive landscape of data platforms.
Sources and Further Readings
[1] Snowflake, March 5, 2024 — Snowflake Cortex LLM Functions Release Notes – — Preview (2024)
[2] Snowflake, Large Language Model (LLM) Functions (Snowflake Cortex) (2024)
[3] Snowflake, Snowflake Puts Industry-Leading Large Language and AI Models in the Hands of All Users with Snowflake Cortex (2024)