Protecting Data Privacy and Confidentiality

Foundation provides powerful data privacy protection through column-level classification and dynamic data masking, ensuring sensitive information is automatically protected based on user permissions while maintaining data utility for authorized use cases.

How It Works

Column-Level Classification: Authorized users can tag individual columns within data products with specific classification and sensitivity levels. For example:

• A customer dataset might have the "email" column marked as PII

• The "customer_id" column marked as Confidential

• The "country" column marked as Public

Dynamic Masking at Compute Time: When users query or access data products, Foundation automatically applies masking based on their role and permissions:

• The data remains unmodified in storage

• Masking occurs dynamically during query execution

• Different users see different versions of the same data based on their access rights

Real-World Example

Consider a sales dataset with customer information:

• Data Analyst (standard permissions): Sees masked email addresses (e.g., j***@company.com), full customer IDs, and complete geographic data

• Marketing Manager (elevated permissions): Sees full email addresses for campaign execution, but social security numbers remain fully masked

• Compliance Officer (audit permissions): Can view all fields unmasked for regulatory reporting

• External Partner with descovery permissions only (restricted access): Sees all data redacted except columns marked as Public.

Why This Approach Matters

Compliance Made Simple

• Automatically enforce GDPR, CCPA, and other privacy regulations

• Maintain audit trails of who accessed sensitive data

Reduced Risk, Maintained Utility

• Minimize data exposure without creating multiple dataset copies

• Enable analytics and ML on sensitive datasets through partial masking

• Prevent accidental data leaks while preserving business value

Operational Efficiency

• No need to create separate "sanitized" versions of datasets

• Classification changes immediately propagate to all consumers

• Single source of truth with multiple privacy-preserving views

This approach ensures organizations can democratize data access while maintaining strict privacy controls, enabling teams to work with sensitive data confidently and compliantly.