Concept card

SAP Knowledge Graph

As of 2026-10-10

Edited by Cédric Mary, sole editor and publication director — 27 years in the SAP ecosystem. How these pages are produced, and where automation is used.

What is SAP Knowledge Graph?

SAP Knowledge Graph is SAP's pre-populated, SAP-managed semantic map of business entities, processes and relationships (SAP Learning material and press coverage of the SAP–Anthropic partnership put the S/4HANA graph alone at roughly 452,000 ABAP tables, 80,000 CDS views and 7.3 million fields — a widely repeated order of magnitude, not an independently confirmed SAP statistic). Since Sapphire 2026 it sits at the core of the SAP Business AI Platform and grounds Joule, Joule Studio agents and BDC. Do not confuse it with the HANA Cloud Knowledge Graph Engine, the RDF/SPARQL triple store you use to build your own graphs.

What it is — and what it is not

SAP Knowledge Graph is a solution that, in SAP's own product wording, connects a company's data fabric to AI for accurate and contextual responses. Concretely it is a machine-readable map of SAP business semantics: which business objects exist (sales order, delivery, billing document, supplier, work center, cost center), what their fields mean, and how they relate across processes and applications. SAP maintains it; customers consume it. The product page lists three families of features: grounding Joule answers in customer data, enabling natural-language exploration and data access for users and agents, and pre-populated knowledge graphs for SAP solutions that are continuously enriched by SAP.

Why it matters

The S/4HANA graph is reported — in SAP Learning material and in press coverage of the SAP–Anthropic partnership, not as an independently confirmed SAP statistic — at roughly 452,000 ABAP tables, 80,000 CDS views and 7.3 million fields: metadata no prompt or naive vector index can carry, whatever the exact count.
Since Sapphire 2026 SAP names the Knowledge Graph as the core business-context component of the SAP Business AI Platform; Joule Studio agents are anchored in it by design.
Mixing up the SAP-managed Knowledge Graph with the HANA Cloud Knowledge Graph Engine leads to wrong scoping: one is SAP content you consume, the other an RDF/SPARQL engine you build on.

Key points

SAP-managed, pre-populated semantic map of SAP business entities, processes and relationships; grounds Joule and agents.
Announced at SAP TechEd, October 2024, for access via SAP Datasphere and Joule in Q1 2025.
S/4HANA graph reportedly built on roughly 452,000 ABAP tables, 80,000 CDS views and 7.3 million fields — a figure repeated in SAP Learning material and press coverage of the SAP–Anthropic partnership, not independently confirmed on a primary SAP product page.
Sapphire 2026: core component of the SAP Business AI Platform; Joule Studio anchored in Knowledge Graph + BDC + SAP Domain Models (GA planned Q3 2026 (past due: no GA announcement found as of 9 October 2026)).
Distinct from the HANA Cloud Knowledge Graph Engine (RDF triple store, SPARQL; GA Q1 2025; metadata-based auto-generation announced for Q1 2026).
KGE interfaces: SPARQL_EXECUTE procedure, SPARQL_TABLE SQL function, SQL_TABLE SPARQL function.
Graphs carry meaning and relationships; facts and aggregations stay in BDC data products, Datasphere or HANA tables.
No public price or SKU for SAP Knowledge Graph — confirm packaging within the Business AI Platform per deal.
SAP Connect keynote (6 October 2026): SAP Knowledge Graph maps more than 7 million data fields, 50 million tables, 50,000+ public and private APIs and 400 data products (SAP's CEO; keynote statement, counting basis not defined, no accuracy results published).

Terms used on this page

SAP Knowledge Graph: SAP-managed, pre-populated graph of SAP business entities, processes and relationships used to ground Joule and agents.
HANA Cloud Knowledge Graph Engine (KGE): Native RDF triple store with a SPARQL engine inside SAP HANA Cloud, for building and querying your own graphs.
RDF triple: W3C data unit subject–predicate–object, e.g. WorkCenter_Z_ASM3 producesProduct Product_C900.
Ontology: Schema of a graph (classes, properties, rules) expressed in RDFS/OWL; validated with SHACL.
SPARQL_TABLE: HANA SQL table function that runs a SPARQL SELECT and returns a relational result set.
SAP Domain Models: SAP models trained on SAP code, metadata and business processes, announced at Sapphire 2026 with GA planned for Q3 2026.
One Domain Model: SAP's harmonised business vocabulary (SalesOrder, Supplier…) carried in BDC data product CSN metadata.

Sources

Full card available to members. What the full card adds: the full decision framework · the SAP vs Snowflake / Databricks / Fabric comparison · the common pitfalls and their fix · the cheat sheet · the architecture schemas · the code blocks · the facts worth quoting.

Open in the app →

Guides that answer with this page

These guides cite this page as one of the sources their answer rests on.

SAP Knowledge Graph

What is SAP Knowledge Graph?

What it is — and what it is not

Why it matters

Key points

Terms used on this page

Sources

Related concepts

Academy modules on this

Deep studies

Watch

Guides that answer with this page