AI is moving quickly and the pressure to show tangible value is mounting. But while investments are soaring and the appetite for adoption is growing, many organisations are finding that their data foundation isn’t quite up to the task of supporting AI agents and applications.

Data lives in different systems on-premises and in multi-cloud environments. Each source has its own structure, logic and business rules. Using that data for analytics, planning or AI often means spending time extracting, cleansing and combining data to rebuild the context for different applications. What does this field mean? Which definition is correct? How does this data connect to that process?

This happens repeatedly across the organisation, creating a so-called “complexity tax” that adds work to make existing data useful. The complexity tax slows down IT teams that are responsible for scale and governance. It also slows down business teams that need faster answers, simpler workflows and confidence in the results.

A business data fabric helps address this challenge by connecting data and business context in a trusted, reusable foundation. That foundation becomes much easier to understand when you see it through the experiences of people using it. Let’s follow three people in an enterprise organisation to see how a business data fabric can support a connected planning workflow from data product creation to planning model generation to final plan updates.

Creating a trusted data product

Our first user is Olga, a line-of-business developer. She has been asked to create a derived data product based on an existing data product from SAP S/4HANA. Her goal is to make the right data available for Ryan, who will use it to create a planning model.

Olga begins in Joule Studio. She describes what she needs in natural language, asking for a simplified cost centre data product that includes only active cost centres, English descriptions and a consolidated list of IDs and descriptions.

The AI agent searches for existing sources and presents options for Olga to review. It does not start from scratch or assume the next step. Instead, it guides Olga through an iterative process, asking for confirmation along the way. Olga can see a preview of the new data product before proceeding, helping her feel confident in the data and the logic because she was involved in the process.

The agent proposes a name for the data product, which Olga then updates to match the naming convention she and Ryan have agreed to use. The agent applies the change and prepares the data product for publishing.

Olga’s experience illustrates how a business data fabric lets business context be reused across systems and domains, which is an important benefit. Olga didn’t need to start with raw data or rebuild logic from scratch. She worked with existing governed data products, while remaining in the loop to create a high-quality data product for the next user.

Turning trusted data into a planning model

The next user in our scenario is Ryan, a planning modeller working in SAP Analytics Cloud. He needs to create an operating expense planning model for Anna, a data analyst responsible for a specific business area.

Ryan also begins his work in Joule, describing the model he wants to create and the level of detail Anna needs. Joule recognises his intent and searches for available data products that match the request. Because Olga’s newly derived data product has already been published and is clearly named, Ryan can easily find it, select it and move forward.

This agent-guided process eliminates the need for Ryan to search available data products, understand their structures, create dimensions and maintain them on his own. The agent builds a model to perform the specific data actions Ryan has requested and does it in seconds rather than hours.

Finally, Ryan asks the agent to add an analytic story on top of the planning model to make it easier for Anna to visualise their results. The Joule agent summarises what it will use, creates visualisations and organises them into an intuitive layout. Ryan fine-tunes the resulting dashboard before sharing it with Anna.

Ryan experienced the second important benefit of a business data fabric: modelling becomes simpler and faster when trusted data sources are readily available. The agent can act on governed data with clear business context and Ryan has confidence in the output.

Planning without tracing every data point

Finally, we meet Anna, a financial planner. She opens the planning model and story Ryan has published. Before making changes, she creates a private working copy so she can adjust her plan without changing the published version.

In Anna’s view, the data is ready to use. Some costs come from SAP S/4HANA. Other data comes from a third-party carrier feed connected through the business data fabric. The data she needs for planning is readily available and Anna does not need to stop and ask where each value originated.

Anna uses natural language to shape the plan, removing certain items and asking Joule to repopulate them using alternative criteria. Joule retains the context from her previous prompts, applies the calculations across dimensions and writes the results back into the plan.

Anna’s experience highlights another benefit of a business data fabric: trust is built into the workflow. The data spans SAP and non-SAP sources, but Anna does not need to verify every data point before moving forward. She can plan with confidence because the data foundation has already done the work of connecting sources, preserving meaning and applying governance.

Three users, one connected workflow

Together, Olga, Ryan and Anna demonstrate what happens when data, context and AI agents work together. In just a few minutes, three people complete work that used to take days and everyone works from the same trusted data foundation.

For IT leaders, the significance is that agentic AI is only as reliable as the data beneath it. A business data fabric gives agents the business context they need to support real work across roles, systems and processes. It helps teams move from fragmented data and repeated manual effort to a simpler way of working, where data is connected, meaning is preserved and trust is part of the process.

Watch the on-demand webinar, Build the data foundation for AI with a business data fabric, to see the demo in action and follow Olga, Ryan and Anna as each step plays out in SAP Business Data Cloud.