# Fabric and Power BI reference path

This is the primary synthetic implementation target because Permadyn Analytics has established Microsoft analytics capability.

1. Land API, database, and secure-file extracts through Fabric Data Factory pipelines into a controlled landing zone.
2. Preserve immutable source batches in OneLake with source timestamps, schema versions, and reconciliation metadata.
3. Transform validated records into a lakehouse or warehouse analytical model using SQL, notebooks, or dataflows according to team skills and workload behavior.
4. Publish shared dimensions and facts to a governed Power BI semantic model.
5. Apply Microsoft Entra ID groups, workspace boundaries, least privilege, row-level security where justified, and audited administrative access.
6. Use deployment pipelines and a PBIPro release record for requirements, metric definitions, review evidence, and release decisions.
7. Monitor pipeline freshness, quality failures, semantic refresh, capacity behavior, and report performance.

Production architecture must validate capacity economics, region, network path, disaster recovery, retention, client tenant ownership, and applicable privacy or contractual controls.