SAP BDC Connect Tutorial: 10 Steps to Connect SAP Business Data Cloud with Databricks

SAP Business Data Cloud (BDC) Connect enables organizations to securely share SAP data with external data platforms such as Databricks, Microsoft Fabric and Snowflake through Delta Sharing technology.

In this guide, we walk through the complete setup process, from creating Data Products in SAP Datasphere to consuming and enriching the data in external platforms and writing results back into SAP Business Data Cloud.

Follow the 10-step tutorial below or watch the accompanying video demonstration at the bottom of this blog.

What is SAP Business Data Cloud Connect?

SAP Business Data Cloud Connect is a capability within SAP Business Data Cloud that enables secure, zero-copy data sharing with external data platforms.

It uses industry standards, with Delta Sharing as a core mechanism, so data stays in place while being accessed by authorized consumers.

Key characteristics:

• Zero-copy access to SAP data using Delta Sharing
• Secure integration with external analytics and AI platforms
• Bi-directional sharing of data products
• Support for an open, multi-platform data strategy

Today, SAP Business Data Cloud Connect integrates with Databricks and is designed to expand to other platforms such as Microsoft Fabric, Snowflake, and Google Cloud.

How it works

BDC Connect flows between three key components: SAP Datasphere, the BDC Cockpit, and your external platform (Databricks, Fabric, …).

Setting up your data on the SAP side:

1. Prerequisite & flow1: make sure that you have provisioned BDC with the correct  systems & formation. Create your Datasphere flow in SAP Datasphere: In SAP Datasphere, build a flow that moves your source data into the object store. This creates the Parquet-based Delta Lake files that serve as the foundation for zero-copy sharing to your external platform.
2. Create a Data Provider Profile: A Data Provider Profile is required before you can create any data products, and is created in the Data Sharing Cockpit in SAP Datasphere. Make sure to set the visibility to Formations so your data products can be consumed across BDC components.
3. Package your objects in a Data Product (DP): Define your Data Product by specifying the artifact space (this is the object store pace of step 1) and required details, then add the tables you want to expose. Set the Switch Status to ‘List’ to make the product discoverable in the catalog — this also triggers the creation of the Delta Share endpoint.
4. Locate the Data Product in the catalog: Navigate to Catalog and Marketplace in your BDC Cockpit, where you can see all active data products and select the one you want to share. Verify the product is listed and the metadata is correctly published.
5. Share it to an external partner: Copy the invitation link from SAP for Me and paste it into the external tool (e.g. Databricks) to establish the secure connection. Once accepted, the external platform is ready to receive your data product via Delta Sharing.

Working with the data on the external platform:

6. Receive the delta-share on your external platform: Active data products become available for consumption once they are mounted to a catalog in the external platform. In Databricks, navigate to Delta Sharing, select the SAP provider, and mount the data product to a catalog.
7. Create your notebook: With the data product mounted, open a new notebook in your external platform and connect it to the shared catalog. The data products can now be used in notebooks, the SQL editor, and AI/ML products.
8. Push the data back to the BDC Cockpit: After enriching or processing the data in your notebook, package the results as a new data product and share it back to SAP BDC. This bi-directional flow lets organizations bring advanced analytics results — such as machine learning predictions — back into BDC for operational use.

Completing the loop:

9. Find the writeback in the shared catalog: Back in the BDC Cockpit, navigate to the shared catalog to locate the data product sent from the external platform. Confirm it has arrived correctly and that the metadata and governance tags are intact.
10. Package your objects in a new Data Product: Finally, wrap the writeback data into a new SAP Data Product within Datasphere. This makes the enriched data available to the rest of your SAP landscape — closing the loop between SAP and your external data platform.

Watch the tutorial video: