Snowflake Cortex

Snowflake Cortex is an intelligent, fully managed service that offers machine learning and AI solutions to Snowflake users. Snowflake Cortex capabilities include:

  • LLM Functions: SQL and Python functions that leverage large language models (LLMs) for understanding, querying, translating, summarizing, and generating free-form text.

  • ML Functions: SQL functions that perform predictive analysis using machine learning to help you gain insights into your structured data and accelerate everyday analytics.

Snowflake Cortex brings state of the art ML and AI solutions to your data, within your Snowflake security perimeter. So workloads that leverage Snowflake Cortex solutions can benefit from Snowflake’s existing security, scalability, and governance capabilities.

The scalable infrastructure behind Snowflake Cortex also powers other intelligent Snowflake features currently in private preview with select customers. Contact your Snowflake account representative to inquire about early access to one of the following features:

  • Document AI

  • Universal Search

  • Snowflake Copilot

LLM Functions

Snowflake Cortex LLM functions use large language models (LLMs) to extract insights from data stored in unstructured and semi-structured formats. Today, the available capabilities include:

  • Perform complex reasoning tasks with your text data using COMPLETE.

  • Extract information from unstructured or semi-structured data using EXTRACT_ANSWER.

  • Automatically get a synopsis of text with SUMMARIZE.

  • Detect the mood or tone of text using SENTIMENT.

  • Translate documents to other languages using TRANSLATE.

Text embedding and vector comparison features, which together enable semantic comparison of text documents, are in private preview with select customers. Contact your Snowflake account representative to inquire about early access to these features.

ML Functions

Snowflake Cortex ML functions provide analysts, data engineers, and business users with powerful machine learning models optimized for scalability and accuracy. We provide the algorithms; all you need is your data. You can access these features through SQL.

  • Forecasting determines which factors in your data influence a value of interest (such as sales) and predicts future values based on those factors.

  • Anomaly Detection finds potential outliers in your data by comparing them to a forecast of the same time period.

  • Contribution Explorer determines the factors that influence a value of interest.

Additional ML features are currently in private preview with select customers. Contact your Snowflake account representative to inquire about early access to these features:

  • Classification function that sorts records into two or more classes based on automatically-detected explanatory factors.

  • Low-code Web interface for using ML functions inside Snowsight without writing SQL.

Snowpark ML

For data scientists and developers, Snowpark ML provides the ability to develop, deploy, and use your machine learning models inside Snowflake using a Python API. When your specific needs can’t be addressed by Snowflake Cortex ML solutions, turn to Snowpark ML to develop your own approach.

Language: English