A data architecture assigns responsibility for source capture, integration, history, computation and consumer access. Integration style, processing topology, storage format and domain ownership can be combined when their contracts remain explicit.
In this chapter
- Kimball, Inmon, Data Vault and hybrid warehouse designChoose integration and presentation models from audit and delivery needs.
- Lambda, Kappa, medallion and hybrid processingSeparate processing/replay, data-quality and domain-ownership decisions.
- Data lake and Delta Lake design: when a warehouse is neededPlace raw retention, catalogs, table commits, compute and analytical serving.