[Snowflake Data Clean Rooms] Row Access Policies

The last big news from Snowflake is how to use row access policies to easily implement distributed Data Clean Rooms. The information is super-fresh and not published anywhere else at this moment, not even by Snowflake.

Rachel Blum, now a Principal Architect and Field CTO Office at Snowflake, has been so kind to meet with us and walk us through this totally new Data Clean Room technique. Rachel is at the heart of everything that has been done on data sharing at Snowflake for the past 2 to 3 years.

Let’s walk through the previous picture to better understand what we are looking for and where the challenge is.

What Our Use Case Is

Most companies collect a huge amount of data today, but they cannot easily share it with their partners, because of legal issues or simply because they cannot give up the raw value of their customer lists. However, you could derive some very precious insights based on the customers you have in common. But how to match those records without having anyone looking at the actual join values? And how to respect restrictions from GDPR, CCPA, or HIPAA, that require you to not expose user information?

This is where the Data Clean Rooms started to play a huge role, and you will later see how deceivingly simple you may implement this…