Why the data stack needs to be re-imagined
Artificial intelligence has rewritten the rules of the game in data and analytics. What used to be primarily reporting and planning now extends deep into operational business. AI models should not only analyze data, but also actively create value – whether by integrating new sources via the Model Context Protocol (MCP) or by intelligently using corporate knowledge with Retrieval Augmented Generation (RAG).
However, this requires an architecture that is agile and scalable – and that operates in the cloud. In a nutshell, the modern data stack is the foundation on which companies will build innovations in the future.
From a monolithic past to a hybrid future
For a long time, SAP was the market leader with its monolithic suites. Everything from a single vendor, seamlessly integrated – but also not very flexible. Today, things look different: companies combine the stability and data quality of SAP with the strengths of specialized providers.
This results in hybrid architectures that combine the best of both worlds: stable SAP data quality on the one hand, and state-of-the-art analysis, AI and sharing functions on the other. This is what makes use cases such as real-time analytics or advanced AI/ML feasible in the first place.
SAP Business Data Cloud: The Center Piece
SAP has recognized this development and adapted its strategy. Since 2023, the company has been regularly integrating partners such as Collibra, Confluent, DataRobot and Databricks into its ecosystem. The result is a multimodal data architecture that brings faster benefits to customers and accelerates innovation.
Since early 2025, the central platform has been the SAP Business Data Cloud (BDC). It bundles data integration, management, reporting, planning, and AI-powered assistants. SAP offers migration options to the cloud – for example, with SAP BW PCE and Datasphere as part of the Business Data Cloud. For companies, this means that existing BW investments can be transferred to the future.
Databricks & Snowflake: The Power Partners
Databricks & Snowflake: The Power PartnersTwo players stand out in this context:
- Snowflake boasts impressive performance in the area of data warehousing, its ability to implement data mesh architectures and, above all, its simplicity in data sharing. Data can be shared along the entire value chain – even with partners who do not have their own Snowflake account.
- Databricks is a leader in data lakehousing, advanced analytics and AI/ML. A milestone was the launch of the SAP Business Data Cloud with Databricks in February 2025. With ‘zero copying’ and ‘delta sharing’, both systems access the same data set directly. For data scientists, this means direct access to structured SAP data in their familiar environment. At the same time, SAP users can easily utilise Databricks’ lakehouse and AI/ML functionalities.
This interaction is a prime example of the new flexibility: no more artificial boundaries between platforms, but an open, seamless data ecosystem.
Conclusion: Fit-for-purpose instead of either/or
The modern data stack is hybrid, cloud-first and fit-for-purpose. It’s not about choosing a manufacturer, but about combining the right building blocks for your own goals.
What is needed for this is a clear data & analytics strategy that is designed across departments – and the courage to explore new approaches. At cbs, we support companies on this journey: from the Data Strategy Pulse Check to the Insight-to-Impact Session to the development of specific use cases.
The first step towards a modern data stack begins right here.
