SAP Business Data Cloud - A Complete Multi-Architecture Platform?

[pbaumann]

SAP announced SAP Business Data Cloud as a multi-data architecture platform integrating SAP technologies like SAP BW, SAP Datasphere and SAP Analytics Cloud. It is building a bridge to Databricks with an integrated SAP Databricks service or a further announced native Databricks integration via Delta Sharing.

While the general approach is to bring SAP data from the huge SAP ecosystem like S/4HANA (ERP), SAP SuccessFactors, SAP Ariba and many more together with non-SAP and Big Data sources and workloads.

Data Architecture Patterns

While many details about the integrated approach are currently discussed, I want to look at the new approach from a data architecture pattern perspective.

Let’s start with a brief overview of the typical data architecture pattern we see in the market:

 The Rise of SAP Business Data Cloud

SAP already partnered with Databricks in 2023 in the context of the Open Data Ecosystem and with leading partners like Confluent, Collibra, DataRobot and Google. SAP did the next step now in February 2025 to not only bring together their on-prem (SAP BW) and cloud data & analytics solutions, they integrate Databricks as SAP Databricks under a unified offering. Additionally a stronger integration with independent Databricks tenants is announced.

But why is this interesting for SAP customers? From a data architecture perspective we can see the chance to cumulate the different data architecture patterns for the best choice. Every pattern introduced before has unique possibilities to integrate, process and deliver data.

• SAP Datasphere is especially strong in the Data Warehouse and (Business) Data Fabric pattern. Now SAP is even completing the Data Fabric pattern by adding more semantics through a knowledge graph and the usage of SAP’s One Domain Model (ODM). This helps to connect data from different sources to a common understanding and support use cases like Self-Service Analytics and AI applications, e. g. Retrival Augmented Generation (RAG) by using LLMs and the new HANA Vector Engine. Additionally the mature SAP BW solution is available as part of SAP Business Data Cloud, and the data can be made available from and for all solutions by generating Data Products from the different solutions on the SAP Data Lake in a Parquet/Delta Lake format. SAP just added the SAP Data Lake, Files, already available for SAP HANA Cloud to enhance the capabilities of SAP Datasphere and supports also Composable Data Platform approach with the 2023 announced Open Data Ecosystem.
• Databricks is initially available as SAP Databricks, a service offering by SAP, seamlessly integrated into SAP Business Data Cloud and able to share data via the Delta Sharing protocol. Databricks is very well known for their Data Lakehouse approach which they even coined in the market. Additionally it can be seen as part of the Modern Data Stack, a Composable Data Platform combining tools as needed in a best-of-breed approach. Additionally Databricks is very flexible to support streaming use cases through differnt components and open and flexible for Data Mesh approaches.

The Multi-Data Architecture Platform

SAP Business Data Cloud forms now a framework delivering all these services in a common and integrated manner. This leads to the following picture, where we see the data architecture patterns each vendor contributes, to a really complete multi-data architecture platform:

Combined, you are able to use the best of booth worlds together in a single platform and adapt the possibilities to your individual needs. You do not have to use every service but always have the chance to expand for new requirements and can easily use a huge amount of capabilities to create value from your data. This flexibility is really interesting in an accelerating data- and AI-driven world.

SAP Business Data Cloud is very new and the time will prove how it will work on the customer side and which new features and degree of integration will be available. Used right, this symbiosis of two vendors can be a really new level for analytical data management and is a clear way forward for many SAP BW and SAP Datasphere customers.

---

What is your experience with SAP Datasphere and Databricks and maybe already with the new SAP Business Data Cloud? Can you leverage the potential of the multi-data architecture approach?

This article is cross-posted at my Substack blog & newsletter.

[Jelena_Perfiljeva]

Nice write-up, Peter! Wow, the data architecture pattern sure have been multiplying lately. 🙂 "Lakehouse"? Blimey...

[pbaumann]

Hi @Jelena_Perfiljeva!

Yes, there are a lot pattern out there. At the same time as we see, todays solutions deliver multiple patterns at the same time. If it comes to Lakehouse currently for the cloud native data platform providers this is the go-to approach. Especially Databricks coined this approach and brings in this pattern into BDC. I tried to visualize this a little bit - first draft: 

The idea is to scale better with cloud resources and build your individual data management system with components you can choose. Databricks brings in an specific approach like Delta Lake as table format which is also the base or the delta sharing/zero-copy approach in BDC. Other providers foster Apache Iceberg or Apache Hudi as alternatives.