Databricks Launches "Clean Rooms" in Public Preview: Secure and Privacy-Safe Collaboration in the Cloud

cloud

As organizations across various industries increasingly rely on data and artificial intelligence (AI) to drive innovation, the need for secure, privacy-preserving collaboration has become paramount. Addressing this challenge, Databricks has launched "Clean Rooms" in public preview, offering a solution that enables seamless and secure collaboration in the cloud while safeguarding sensitive data.

A Secure and Flexible Collaboration Environment

Many organizations face significant challenges when collaborating, often having to relinquish control over how their sensitive data is shared, leading to risks of misuse and privacy breaches. Databricks Clean Rooms offers a solution that allows organizations to collaborate effectively and securely without compromising data protection.

Powered by Delta Sharing, Databricks Clean Rooms enables businesses to collaborate with partners and customers across any cloud without exposing sensitive data. Participants can securely share and combine their existing data and run complex workloads using any programming language, such as Python, which offers native support for machine learning (ML). Importantly, data remains in place, ensuring that each organization retains full control over how it is used.

Data-Driven Innovation Across Industries

Databricks Clean Rooms is designed for enterprises seeking to accelerate innovation through data-driven insights. The solution supports a wide range of analytics and AI workloads, from simple SQL queries to complex computations in Python. This flexibility is crucial for industries such as advertising, retail, manufacturing, healthcare, and financial services, where collaborative data analysis can provide significant competitive advantages.

For instance, in the advertising industry, Clean Rooms allow advertisers and publishers to analyze campaign performance without compromising user privacy. In healthcare, researchers can collaborate on patient data analysis to develop new treatments while maintaining confidentiality.

Scalable and Cross-Platform Collaboration

One of the standout features of Databricks Clean Rooms is its ability to facilitate collaboration across different regions, clouds, and platforms without the need for data replication. Powered by Delta Sharing, this secure and open collaboration enables partners to work together on data and AI models, including non-tabular and unstructured data, while ensuring the privacy of underlying data.

Databricks is also developing new features, such as "Sharing for Lakehouse Federation," to further expand the possibilities for secure collaboration across data platforms.

Easy Setup and Usage

Despite its powerful capabilities, Databricks Clean Rooms is easy to set up and use. Users can create a secure environment by selecting their preferred cloud provider and region, then invite collaborators to securely share data. Analysis takes place in an isolated environment, ensuring that no participant can directly access another's data.

The solution also allows collaborators to share notebooks containing approved code, which can be executed within the secure environment using serverless compute, facilitating data analysis in a controlled and private manner.

Use Cases and Availability

Databricks Clean Rooms is already being utilized in sectors such as advertising, retail, manufacturing, healthcare, and financial services, providing solutions ranging from predictive modeling to fraud detection and personalized financial products.

The public preview of Databricks Clean Rooms is available on AWS and Azure, and its adoption is expected to grow as more organizations seek secure ways to collaborate on data and AI projects. Interested organizations can request access to the public preview through Databricks.

In summary, Databricks Clean Rooms offers a robust solution for secure and privacy-safe collaboration in the cloud, enabling organizations to innovate and harness the power of data and AI without compromising privacy.