Challenge
KPIs repeatedly required complex joins across 15+ SAP source tables.
SAP-Driven Data Warehouse Build
Reusable data entities instead of repeated SAP table logic.
A Bronze → Silver → Gold warehouse approach that consolidates source complexity into shared business entities and governed KPI marts.
Delivered
Reusable Bronze, Silver and Gold layers with business entities and quality controls.
Enabled
Documented KPI marts and reusable data structures for downstream analytics.
Challenge
Sales documents in SAP are spread across 15+ tables (VBAK, VBUP, VBKD, VBPA, VBAP, and more). Each KPI required complex joins - error-prone, CPU-heavy, and frequently re-implemented across teams. Data quality issues surfaced in reports rather than at ingestion. Onboarding a new use case meant re-deriving the same business entities every time.
Approach
We built a structured DWH using a Bronze → Silver → Gold pattern. Bronze layer holds raw SAP extracts. Silver consolidates sales documents, deliveries, and billing into reusable business entities with quality controls and incremental updates: the joins that depend only on the SAP architecture, not on business logic, are done once here, together with data cleaning and validation. Regular jobs load only created or modified records, and full or partial recovery is also possible. Gold layer applies KPI-specific logic and serves data marts (CSL, Freshness, etc.) ready for reporting and AI use cases.
The approach is systematic: each source domain is built once, harmonized once, and reused across analytics workloads.
Architecture and reusable data layers
The project's SAP-to-warehouse diagram.
Business entities beyond sales documents
The same approach builds other Silver-layer business entities from their SAP source tables. Which entities apply depends on the SAP modules in use; these are examples from the project documentation.
| Business entity (Silver) | SAP tables (Bronze) |
|---|---|
| Controlling Documents | COEP, COBK |
| Accounting Documents | BSEG, BKPF |
| Production Orders | AFKO, AFPO, AUFK |
| Sales Documents | VBAK, VBAP, VBKD, VBPA, VBUK, VBUP |
| Purchasing Documents | EKKO, EKPO, EKPV |
| Purchasing Document Account Assignment | EKKN, EKKO, EKPO |
| MD Classes/Characteristics | KSSK, KLAH, TCLA, CABN |
From SAP source tables to KPI data marts
-
01
Bronze
Raw SAP extracts.
-
02
Silver
Sales documents, deliveries and billing consolidated into reusable business entities, with quality controls and incremental updates.
-
03
Gold
KPI-specific logic and data marts (CSL, Freshness) ready for reporting and AI use cases.
This project uses SAP as its source system. The Bronze/Silver/Gold approach itself is not SAP-specific.
Outcomes
Qualitative project outcomes; no measured performance or cost figures are claimed.
Streamlined SAP integration across 15+ source tables
Reusable Silver entities (Sales Documents, Deliveries, Billing)
Stable performance and reduced ad-hoc joining
Governance and quality controls built into pipelines
Faster onboarding of new use cases and new sources
Foundation for downstream analytics and AI
Technology
SAP
DWH platform
ETL
Data quality framework
Solution areas: Data Foundations
