Privacy-Compliant Data Aggregation

Design privacy-compliant aggregation for {{sensitive_data}} under {{regulation}}:

1. Regulatory requirements:
   - What PII is in scope? (name, email, IP, location, etc.)
   - Minimum aggregation level to comply?
   - Retention policies?

2. Aggregation strategy:
   - Suppress low-count cells (k-anonymity: min group size {{min_group_size}})
   - Generalize quasi-identifiers (e.g., zip code → region)
   - Hash or tokenize identifiers before analysis
   - Differential privacy noise (if needed)

3. Practical implementation:
   - SQL / dbt logic for safe aggregates
   - QA checks to prevent de-anonymization
   - Access controls (who can query raw vs aggregated data?)

4. Trade-offs:
   - Utility loss from aggregation (coarser granularity)
   - Risk mitigation: what re-identification attacks are we protecting against?

5. Documentation:
   - Data governance policy
   - Audit log for data access

Deliverables: aggregation rules + SQL implementation + compliance checklist.