Data Sharing and Collaboration

Secure data sharing leverages Snowflake's architecture where compute and storage are separated, allowing consumer accounts to query shared data directly using their own compute resources while data remains in the provider's account. This eliminates data duplication, ensures consumers always access current data, reduces security risks from data proliferation, and simplifies distribution.

Providers create shares containing references to database objects (tables, secure views, secure UDFs). Secure views enable personalized data views within shared datasets by filtering rows or masking columns based on consumer identity. Consumers create databases from shares, paying only for their own compute consumption.

Collaboration Extensions

• Snowflake Marketplace for discovering and acquiring external datasets
• Data Clean Rooms for multi-party analysis without exposing raw data
• Collaboration objects enabling bidirectional data interaction
• Commercial listings with subscription or usage-based pricing