Living documentation that stays in sync with engineering, cloud, AI, and compliance activity. Documents update themselves as the organization changes.
Governs how data is sourced, minimized, tracked, and protected — including PHI — for AI development.
Ensure data used in AI systems is lawful, minimized, traceable, and safe.
All datasets used for training, fine-tuning, evaluation, or retrieval.
| Activity | R | A | C | I | Cadence |
|---|---|---|---|---|---|
| Approve data source | Data Governance Lead | GRC Lead | Legal | Team | Per source |
| Register lineage | Data Engineer | Data Governance Lead | AI Platform | Team | Continuous |
| Approve PHI use | GRC Lead | CISO | Legal, Clinical | Team | Per dataset |
Inputs
Outputs
| Metric | Target |
|---|---|
| PHI datasets with approved basis | 100% |
| Datasets with complete lineage | ≥ 98% |
| Record | Retention |
|---|---|
| Data source records | 10 years |
| Lineage | Life of dataset + 5 years |
See the Versions tab for the full change history maintained by the Auto-Doc Engine.
Document code: SOP-AIDLC-02 · Aligned to HIPAA §164.502(b) / GDPR Art. 5, 25 / NIST AI RMF Map. Controlled document — reproduction outside the UBC QMS requires the Quality Manager's approval.