From ERP silos to the AI Data Cloud: How SAP and Snowflake can future-proof your data strategy

Many organisations face the challenge of combining process-related insights from their SAP systems with the agility and scalability of a modern cloud platform. Thanks to the close partnership between SAP and Snowflake, the SAP Business Data Cloud (SAP BDC) and the Snowflake AI Data Cloud are merging to form an integrated data and AI ecosystem.

SAP BDC is SAP’s cloud-based platform for the centralised management and delivery of business data from the entire SAP system landscape, from S/4HANA to BTP. Snowflake is a scalable AI Data Cloud capable of processing structured and unstructured data from any source and delivering it across the enterprise.

In future, the replication of SAP data will be reduced through deep integration, thereby paving the way for true data democratisation across the entire organisation. This technological bridge enables self-service analytics at the touch of a button, allowing business departments such as Controlling to make decisions based on a consolidated, AI-powered data foundation without having to go through the IT department.

The challenge: the ‘silo dilemma’ in enterprise IT

For decades, SAP has been the backbone of the global economy. A company’s most valuable information lies within the modules for Finance (FI), Controlling (CO) and Supply Chain Management (SCM). However, this data has evolved over time, is highly standardised and is often difficult for modern analytics tools to access. Anyone wishing to use this data in a cloud environment such as Snowflake faces enormous obstacles:

• Manual data replication: Data must be extracted via complex ETL processes and physically copied into other systems.
• Loss of context: The semantic logic of SAP tables is often lost during extraction. A technical field such as ‘WRBTR’ in the ‘BSEG’ table is simply not self-explanatory.
• High latency: Batch processing means that analyses are often based on data that is already hours or days old.

The new strategic partnership between SAP and Snowflake addresses precisely these pain points and creates an integrated solution that goes far beyond a simple interface.

The technological revolution: zero-copy data sharing

At the heart of this new collaboration is the shift away from rigid connectors towards dynamic integration. Instead of moving data from A to B, SAP and Snowflake utilise native zero-copy integration (see Fig. 1). This process follows a clear, highly efficient logic that minimises data transfer and maximises performance:

Deployment in SAP Datasphere: Within SAP Datasphere, the complex raw data from the SAP systems (S/4HANA, BW/4HANA) is harmonised and refined into data products. These contain not only the raw data records, but also the complete business logic and semantic definitions.

Sharing via SAP BDC: These data products are made available to Snowflake via the SAP Business Data Cloud (BDC) service. The data is no longer physically moved; instead, a secure, bidirectional connection is established, enabling real-time access.

Integration via Horizon Catalog and Iceberg: In Snowflake, the shared BDC share is integrated into the Snowflake Horizon Catalog via a Linked Catalog. The SAP data products are mapped as Iceberg tables without any physical copying of data. This is where the true strength of the zero-copy approach becomes apparent. Snowflake merely registers the metadata, whilst the original data remains at SAP in its optimised storage format. Nevertheless, the tables can be queried directly in Snowflake as if they were stored locally.

Semantic views: Based on these Iceberg tables, semantic views are created in Snowflake that mirror the data products from SAP on a one-to-one basis. The key point is that the business logic – i.e. dimensions, hierarchies and attribute descriptions – is automatically imported from SAP. This eliminates the need for manual remodelling. For the user, this means they can work in Snowflake using the familiar business logic, whilst the system fully utilises the performance benefits of the cloud architecture in the background. At the same time, the views serve as the basis for Cortex Analyst, enabling the data to be queried later using natural language.

Specifically, two product variants are available: SAP Snowflake and SAP BDC Connect for Snowflake, both of which are built on the same technical foundation and offer the same functionality. Whilst SAP Snowflake is offered as a Solution Extension within SAP BDC for new Snowflake customers, SAP BDC Connect for Snowflake is suitable for customers who are already using Snowflake.

Preservation of semantic intelligence

A key advantage over older methods is the preservation of the business context. Thanks to the seamless integration of SAP data products into Snowflake, the metadata, hierarchies and business logic defined in SAP Datasphere flow directly into the semantic views. This means that business departments in Snowflake do not work with raw data sets, but with validated business objects such as sales orders or customer master data. Misinterpretations are prevented and analyses can be carried out significantly faster. As a result, the time required for data engineering is drastically reduced, whilst data quality improves.

Security and governance through Snowflake Horizon

Alongside content intelligence, the Snowflake Horizon Catalogue ensures the necessary governance. Once the SAP data products are registered as iceberg structures in the Catalogue, they are subject to the Snowflake platform’s central security policies. Administrators can precisely control access down to the row and column level without placing a burden on the operational SAP source systems. This seamless monitoring of data provenance and access control ensures that even highly sensitive financial or HR data is processed in compliance at all times, whilst data sovereignty is consistently maintained through the zero-copy approach. As no copies of the data are in circulation, the risk of data leaks is reduced and compliance with data protection regulations is significantly simplified.

Data integration in practice: Analysing SAP and non-SAP data seamlessly with Snowflake Cortex

Another strength of the partnership becomes apparent where different data environments converge. In many organisations, data from the central SAP system must be combined with information from third-party systems – such as CRM platforms or local ERP solutions used by subsidiaries – to gain a complete overview. Until now, this consolidation required in-depth technical expertise and lengthy coordination processes with the IT department. The integration of SAP data products into the Snowflake environment removes this hurdle. As the data is already harmonised and available via the Catalog Service, it can be linked to non-SAP sources without the need for technical expertise.

During the critical phase of month-end closing, this approach enables a whole new level of speed. Instead of manually consolidating data sets, a controller can use Snowflake Cortex Analyst to access the data sets authorised for them directly. The controller does not need to be proficient in SQL to run complex queries across system boundaries. Using natural language, complex analyses across millions of data records can be carried out within seconds. The AI acts as a tool that bridges the gap between the systems. It identifies the relevant KPIs in the SAP and non-SAP tables, combines them in line with business logic and calculates the required metrics on the fly.

This process delivers a decisive advantage for corporate management. Instead of waiting days for processed reports, the Controlling department can provide management with validated KPIs in real time. The time between the generation of data in day-to-day operations and the strategic decision is reduced to a minimum. As the analysis takes place directly on the consolidated data set in Snowflake, the reliability of the figures is maintained, whilst flexibility in the evaluation increases significantly.

Conclusion: Agility through a unified data foundation

The partnership between SAP and Snowflake is the technological response to the growing demand for greater speed and self-service analytics. Combining SAP’s process stability with Snowflake’s innovative strength creates a foundation that goes far beyond traditional reporting.

Furthermore, this approach promotes genuine data democratisation, as business users can independently gain in-depth insights, whilst the IT department is noticeably relieved of the burden as complex export and transformation processes are eliminated. Those who connect these two worlds today secure a decisive competitive advantage in the age of data-driven decisions and artificial intelligence.

Would you like to bridge the gap between your SAP system and Snowflake? At adesso SE, we support you with two specialised teams: our SAP Analytics Team knows your SAP landscape inside out – from the data models in S/4HANA, through the semantic layer in SAP Datasphere, to deployment in SAP BDC. Our Snowflake team has the expertise to deliver your data products in Snowflake with high performance and make them usable for AI applications. Together, we cover the entire integration process – from the initial assessment of your SAP data landscape right through to the productive use of Cortex AI on your SAP data in Snowflake. Get in touch with us.