Imagine building your own personalized California wine country visit assistant based on a private, structured dataset and a RAG-based approach to Gen AI.

This video walks you through the process step-by-step.
First, you’ll learn how to use Fivetran’s automated data movement platform to replicate a custom, structured winery and vineyard visit dataset (wineries and vineyards across all California wine country regions) from a relational database source to the Snowflake Data Cloud.

You’ll then see how Fivetran’s automation allows you to move data without the hassle of schema creation or schema management, achieve reliable incremental syncs and change data capture, and handle a range of data privacy requirements.

From there, you’ll be walked through creating LLM-friendly data transformations in Snowflake for this structured dataset and how you can use the newly transformed data and Snowflake Cortex to build an interactive California Wine Country Visit Assistant application in a Snowflake Streamlit native application.

You’ll see how the application generates dynamic, context-aware responses about wineries and vineyards, leveraging a RAG-based approach and vector similarity searches for enhanced information retrieval. Plus, you’ll see how this approach enables the creation of a customized, detailed wine country visit itinerary using the private, custom dataset.