Snowflake Native Vector type support and Snowflake Cortex Search support

We’ll add the ability to write into Snowflake tables using the new “VECTOR” type in a column. To do so, a new Cortex Embed component will be available, giving the ability to select the embedding model to use as well as the size of the embedding. This will enable the loading of context to create a native RAG pipeline on Snowflake.

The other part of this story will be a new component to integrate to Snowflake Cortex Search, and give the ability to retrieve an embedding. This will help data engineers design pushdown RAG pipelines using Snowflake Cortex capabilities.