Data layer

The Data Layer governs how data is stored, accessed, and used under enforceable consent constraints. It acts as the boundary between raw data domains and protocol-level computation.

Its responsibilities include:

Consent Binding and Enforcement
All data is bound to one or more Consent-as-Code (CCL) policies at ingestion. Every access request must satisfy constraints such as purpose, geography, retention, k-anonymity, and privacy budget (ε) before execution.

Data Residency and Control
Data remains within its original domain, isolated by tenant and geography. Ownership of encryption keys is retained by data providers, with support for rotation and revocation.

Identity Abstraction (AudienceID)
Identity signals are combined within trusted environments to produce a stable but irreversible audience identifier, enabling deduplication and frequency control without exposing raw identities.

Privacy-Preserving Outputs
All outputs are generated through ESQ and must satisfy k-anonymity and ε constraints. Only aggregated, compliant audience segments are returned—never row-level data.

Auditability and Lineage
Every data access and transformation is recorded as a structured audit trail, enabling deterministic replay in audit environments without exposing sensitive data.

This layer ensures that data usage remains permissioned, auditable, and constrained by enforceable privacy guarantees.