Question

A data engineering team wants to build a transformation pipeline in Python that manipulates large tables using DataFrame operations. A key requirement is that the data processing logic must execute directly on Snowflake's virtual warehouses (pushed down to the compute layer) rather than pulling data out to an external client machine. Which Snowflake feature is designed to meet this requirement?

Accepted Answer

Snowpark, which lets developers write DataFrame code in Python that is translated into SQL and executed on Snowflake compute

Answer

Streamlit apps, which render interactive dashboards from the results of SQL queries

Answer

Snowflake Cortex, which provides serverless LLM functions for text summarization and translation

Answer

The Snowflake Native App Framework, which packages applications for distribution through the Marketplace