Route equivalence
The same source fixture produces the same Bronze meaning through events, files, APIs or CDC.
The lakehouse separates arrival, preservation, reconciliation and business meaning so a transport choice, pipeline convenience or dashboard cannot silently become truth.
Ingestion preserves and reconciles declared source state. Joins, aggregates, entity resolution and measures begin only after the governed consume boundary.
The rule prevents transport logic from acquiring business authority and keeps replay possible from evidence whose meaning has not been rewritten.
The medallion is useful only when each layer owns a distinct class of change and authority.
| Layer | Question | Allowed change | Not allowed |
|---|---|---|---|
| Landed | What arrived? | Transport evidence and object identity | Rewriting the source object |
| Bronze | What did the source state? | Route-independent envelope and source payload | Business interpretation |
| Silver | What is the declared current and historical entity? | Typing, keys, sequence, CDC and SCD behaviour | Joins and measures |
| Gold | What does the business entity mean? | Conformed dimensions, facts and grain | Presentation-specific definitions |
| Semantic | How is the shared measure evaluated? | Named measures and governed dimensions | A second calculation inside a dashboard or agent |
Classification, masking, tokenisation, lineage and freshness are reviewed at the governed object where all consumers inherit them.
The same source fixture produces the same Bronze meaning through events, files, APIs or CDC.
Keys, sequence and SCD type are declarations with fixtures for late, duplicate and out-of-order change.
A measure is declared once. Dashboards and agents query it rather than carrying parallel formulas.
Classification and policy travel with data. Tokenised or masked surfaces are deliberate products, not ad-hoc query behaviour.
An objective names the source, expected cadence, evidence, owner and response—not only a green tile.
Derived targets can be rebuilt from preserved evidence in isolation, compared, and promoted through the same acceptance path.
A structurally valid model can still encode the wrong business meaning. Named data owners must accept grain, measures and exception policy.