Introduction to Snowflake Data Clean Rooms¶

About data clean rooms¶

Data clean rooms offer a secure way to gain valuable insights while protecting sensitive information. They allow you to combine and analyze data from different parties without worrying about the privacy concerns that go with sharing raw data. With data clean rooms, multiple parties can collaborate without revealing their underlying data.

Benefits of data clean rooms include:

• Enhanced privacy — Protects sensitive data while enabling collaboration.

• Deeper insights — Combines data from multiple sources for richer analysis.

• Increased security — Reduces the risk of unauthorized access.

How Snowflake Data Clean Rooms work¶

With Snowflake Data Clean Rooms, all analyses are conducted within the secure environment of the clean room. Collaborators are able to return aggregated results and insights, but cannot directly query the raw data in the clean room. The collaborator who is sharing their data can define what analyses are available to the other collaborators, allowing them to tightly control how their data is used.

A Snowflake Data Clean Room also uses privacy-enhancing techniques on its data such as:

• Using differential privacy to add noise to results in order to prevent someone from identifying whether a particular individual is in the data.

• Encrypting the data, then running multi-party computations directly on the encrypted data.

Clean room collaborators¶

Snowflake Data Clean Rooms use the concept of a provider and consumer, similar to other Snowflake features like Secure Data Sharing. The data owner is a provider who uses a clean room to safely share data with a consumer. The consumer installs the clean room in their own account and analyzes data in the clean room, including joining their own data with the data of the provider.

Tasks associated with clean room collaborators include:

Provider:

• Create a clean room.

• Add data to a clean room.

• Configure a clean room to control how a consumer can interact with data.

• Share a clean room with a consumer.

Consumer:

• Install a clean room.

• Add datasets to the clean room.

• Analyze data in the clean room, including joining consumer data with the provider’s data.

Within the clean room environment associated with a Snowflake account, a collaborator can be the provider of one clean room while acting as the consumer of another.

For information about adding a collaborator to your clean room environment, see Add collaborators.

Collaborators who are not Snowflake customers¶

The provider of a clean room must have a Snowflake account. However, a provider can collaborate with consumers who do not have a Snowflake account by creating a clean room managed account for the consumer. To begin collaborating in a clean room, the consumer simply accepts the provider’s invitation to use the managed account.

A managed account has the following characteristics:

• Exists in the same cloud platform and region as the Snowflake account of the provider who created the managed account.

• Must exist in the same cloud platform as the consumer’s data. As a result, the Snowflake account of the provider must exist in the same cloud platform as the consumer’s data.

• Does not behave the same as a Snowflake reader account. The consumer does not access the managed account outside the context of the clean room environment.

• Can be converted to a Snowflake account in the consumer’s organization if they want to become a Snowflake customer.

For a description of the tasks related to managed accounts that providers and consumers perform, see Work with a clean room managed account.

Working with a Snowflake Data Clean Room¶

A Snowflake Data Clean Room is designed for both business and technical users. You have two options for working with a clean room:

• Web app — An easy-to-use interface that makes privacy-enhanced data collaboration accessible to a wide base of users, including non-technical business users. Collaborators can use pre-defined analysis templates including audience overlap, reach and frequency, and last touch attribution. For an overview, see Snowflake Data Clean Rooms: Web app overview.

• Developer edition — A complete set of APIs that allow a technical audience to work with clean rooms programmatically, including the ability to build custom applications and to customize analysis templates and ML models. For an overview, see Snowflake Data Clean Rooms: Developer edition overview.

Next Step¶

Before users can work with the web app or developer edition, a clean room administrator needs to sign up for the clean room environment and configure the Snowflake account associated with it. For details, see Getting started with Snowflake Data Clean Rooms.