Blog Detail

SAP plans to deliver hundreds of new data products by the end of 2025, making it possible for developers to build intelligent applications, the company revealed at last week’s Sapphire conference in Orlando.

Data products are composed of “curated and governed data, metadata and business semantics from SAP applications and third-party sources,” the company noted in its SAP Sapphire Innovation Guide 2025.

The data products will be offered through Business Data Cloud (BDC), which is SAP’s new managed data service that incorporates Datasphere and SAP Analytics Cloud (SAC).

A Modern Twist on the Data Lake

This comes at a time when very few business or technology leaders trust their data, Irfan Khan, SAP’s president and chief product officer in SAP Data & Analytics, told The New Stack. According to a SAP survey of approximately 1,200 respondents, only 34% trust their own data, he told us.

“So a very, very low percentage in comparison to the huge investment that customers put into managing, curating, preparing their data,” he said.

Essentially, the BDC is a modern data lakehouse architecture, but with all the integration work done by SAP, explained Khan. The BDC officially launched on February 13, 2025, as part of SAP’s “Business Unleashed” event. It’s more of a Software as a Service (SaaS) play than a storage move because it provides governance and manages the data coming out of the source SAP application, he said.

“You minimize the discrepancy between the quality of the data because we serve it up,” he said.

A data product, then, is a defined foundation of data coming out of the core ERP applications.

“We’ve essentially stitched together the semantical model that goes across all of the different lines of business, then we declare these data products coming out of the individual source systems, and we harmonize them into one single domain model for SAP,” he said. “This really is a technical lift and shift where SAP took that responsibility on.”

The Benefit to Developers

Previously, customers built the integrations themselves. Now, SAP makes the data product accessible to developers, who can use these offerings to build intelligent applications, he said.

Customers previously had to handle the data integration via extract, transform and load (ETL) or custom pipelines to pull the data out and put it in some sort of defined environment, typically a dashboard.

It’s different than data as a service, which Khan said implies the data is raw. ERP data tends to be highly curated, very semantically enriched and contextual.

On the ERP side, there’s an Advanced Business Application Programming (ABAP) push framework, SAP’s proprietary, high-level language. The BDC uses the ABAP push framework, which constantly monitors the customer’s estate. So if a customer adds a new sales order, SAP will have the necessary data products that represent a sales order as an entity, and then that will be persisted inside of BDC.

This means SAP developers can build the apps on BDC rather than building natively, which would require integrating the underlying source of SAP applications and non-SAP applications, he explained.

Build, SAP’s low-code development platform, becomes the frontend for “pro-code development,” he said.

“If you want to develop a pro-code application that consumes data of SAP, you can now essentially consume data products natively out of BDC, as well as these intelligent applications, which are also consuming the same data products,” he said. “Data products can be used in a variety of different use cases, including Build, with the consumption you gain out of that.”

Getting It Together for AI

“If you look back over SAP history, we’ve had a data strategy before,” Khan said. “It was HANA, once upon a time in the Datasphere, but with the advent of AI and agentic, we have to really get our act together to help customers accelerate the time to market and time to value.”

Hence, SAP sees BDC and business AI as two sides of the same coin, he added.

Before the BDC offering, SAP customers who wanted to use ERP data for forecasting would have to get the data out of the ERP, build a dashboard to consume the data, build a model, build a queue and then run a simulation on top of all that, he said.

BDC incorporates the SAC, which uses Compass for simulations, specifically Monte Carlo simulations. That can be coupled with AI.

There’s also a knowledge graph, which is a structured representation of knowledge that models real-world entities — people, places, things, events, concepts — and the relationships between them.

“We can hydrate a knowledge graph, and you can start asking those very open-ended questions: Which of my sellers in the east are currently on path to make quota, but have a huge gap to plan for 2026 because I haven’t built enough pipeline?” he said. “Now you can have a large or small language model that covers the request and have it hydrated with tabular SAP data.”

SAP also offers support for vectoring via HANA Cloud. Vectoring supports AI functions such as semantic searches by comparing the vector of a query to the vectors of stored data.

All this gives customers something they were missing previously with data efforts, according to Khan: business context.

“It’s a game changer,” Khan said. “Most customers now have the most important ingredient that they missed before, which is a business context. That business context is a culmination of BDC and business AI in that strong coalition, coming together.”

The post SAP Simplifies ERP Data Access for Developers appeared first on The New Stack.