Placeholder image

Fabric IQ access to SAP Business Data Graph

| Mario De Felipe |

Fabric Graph Fabric IQ

YouTube

Spotify

Apple

Episode #297

Introduction

In episode 297 of our SAP on Azure video podcast we talk about Fabric IQ access to SAP Graph.

AI agents are becoming more and more powerful, and in the SAP world one of the big questions is still: how do we give these agents meaningful access to SAP business context without just copying data around? This is where things like SAP Graph, SAP ontologies, knowledge graphs, MCP servers and Fabric IQ become really interesting.

Mario has been doing a lot of work in this space. The MCP Server for SAP Datasphere environments that he has developed has gained a lot of popularity and is used by several customers.

So for today I am really happy to have Mario de Felipe joining us for the first time, to show us what he has built and to talk about how all of this can help bring SAP context into Microsoft Agents.

Find all the links mentioned here: https://www.saponazurepodcast.de/episode297

Reach out to us for any feedback / questions:

#Microsoft #SAP #Azure #SAPonAzure #FabricIQ #Fabric #Copilot

Summary created by AI

  • Integrating SAP Business Context with Microsoft AI Agents:
  • Holger, Mario, and Goran discussed how Mario’s MCP server for SAP Datasphere enables Microsoft AI agents to access SAP business context without replicating data, leveraging SAP Graph, ontologies, and Microsoft Fabric IQ to unify and reason over distributed enterprise data.
    • MCP Server Overview: Mario explained that the MCP server for SAP Datasphere provides BI teams with view-only access to SAP data via published APIs, allowing agents to query and understand data structures without modifying records. The server is widely adopted and supports integration with Microsoft Copilot and other agents.
    • Challenges of Multi-System SAP Environments: Mario described the complexity of SAP landscapes where entities like suppliers exist across multiple systems (e.g., S/4HANA, Ariba), requiring agents to reason across sources and unify identities and semantics, rather than relying on data replication.
    • Role of Ontologies and Knowledge Graphs: The team discussed how ontologies and knowledge graphs provide a semantic layer that defines business entities, relationships, and rules, enabling AI agents to interpret and answer complex business questions by grounding queries in business meaning.
    • Microsoft Fabric IQ and SAP Integration: Mario outlined how Microsoft Fabric IQ allows customers to build ontologies and digital twins, and how these can be connected to SAP’s API Composition (Graph) to provide live, unified business context for AI agents, without replicating SAP data into Microsoft platforms.
    • Business Value and Future Directions: The participants highlighted that this integration paradigm is new and evolving, with Fabric IQ still in preview and both Microsoft and SAP expected to add features, including automation of graph creation and support for additional data sources.
  • Demonstration of Live SAP Data Access via Fabric IQ Agents:
  • Mario demonstrated how an AI agent in Microsoft Fabric IQ can answer business questions by leveraging a live connection to SAP’s business data graph, showing unified supplier and customer information across S/4HANA and Ariba without data replication.
    • Manual Graph Construction in SAP Integration Suite: Mario showed how the business data graph is manually built in SAP Integration Suite by adding data sources and defining relationships, with the current process requiring user interface interactions but expected to be automated in the future.
    • Ontology Synchronization Between SAP and Microsoft: The demo illustrated that the ontology created in SAP Integration Suite is ingested into Microsoft Fabric IQ, allowing agents to reason over the combined business context from both platforms in real time.
    • Supplier Query Example: Mario queried the agent about which supplier provides a specific material for a blocked order, and the agent performed two calls: first to the Fabric ontology to understand the entity types, then to the SAP business data graph to retrieve live supplier information from both S/4HANA and Ariba.
    • Customer Credit Status Query: A subsequent demo question asked whether a customer is blocked and their credit exposure; the agent distinguished between order, delivery, and credit blocks, providing a detailed, grounded answer by referencing the SAP ontology and live data.
    • No Data Replication Required: Mario emphasized that all agent responses are based on live SAP data accessed via the business data graph and ontology, with no need to replicate or synchronize data into Microsoft systems, ensuring up-to-date and consistent answers.
  • Limitations and Future Enhancements for Fabric IQ and SAP Integration:
  • Mario and Holger discussed current limitations of Fabric IQ, such as the requirement for data to reside in OneLake and manual graph creation, and outlined anticipated enhancements including support for delta tables and broader automation.
    • Current Platform Limitations: Mario noted that Fabric IQ currently requires data to be in OneLake and does not yet support delta tables, limiting integration scenarios with other data products like BDC.
    • Expected Feature Additions: Future updates are expected to enable delta table support and automate graph and ontology creation, expanding the range of data sources and reducing manual setup.

Links

Video index

Based on and using Hugo Cookbook Theme, Font Awesome Icons and Bulma CSS. | Thanks to Recipes4AJ to get me started. | Contact me via