des-governance-and-security

Data governance controls are established by defining dataset ownership, access policies, PII classification, retention enforcement, and audit logging. This process requires complete dataset inventories and data contract inputs to ensure release readiness for sensitive data products. HALT checkpoints validate missing upstream evidence.

Implementing RBAC and row-level security involves specifying role-based access plus column-level isolation so each consumer role queries only authorized data. This access control and isolation strategy ensures safe querying across shared metrics and cross-team stewardship scenarios involving regulated data.

PII classification is defined by assigning sensitivity levels to datasets and selecting masking, tokenization, or encryption strategies per PII column. This privacy treatment ensures regulated data is appropriately protected before production release.

Retention and deletion procedures are enforced by defining retention windows, documenting right-to-be-forgotten flows, and ensuring audit logging for query and export actions. This enforces regulatory obligations and compliance for data products with sensitive information.

Explicit data governance controls are needed when preparing Silver, Gold, or Serving datasets for production release. This is especially required where PII, regulated data, cross-team ownership, external access, or compliance obligations apply to the data product.

Release readiness guardrails use HALT stop conditions to halt the process when required upstream evidence, ownership, regulatory scope, or security decisions are missing. This prevents data-platform trust and compliance failures by blocking incomplete governance artifacts from production.