To combine data from Azure Databricks and an existing Fabric lakehouse using data pipelines for Databricks data and shortcuts for the lakehouse, which Fabric artifact should you create?

To combine data from Azure Databricks and an existing Fabric lakehouse using data pipelines for Databricks data and shortcuts for the lakehouse, which Fabric artifact should you create?

• A. a data lake
• B. a lakehouse
• C. a dataflow
• D. a dataset

Answer: (B)

The central idea is that a lakehouse provides the unified, governed layer you need to bring together multiple data sources. By using data pipelines to bring in Databricks data and shortcuts to reference the existing Fabric lakehouse data, you can create a single lakehouse that serves as the integrated foundation for querying and analysis. The lakehouse acts as the collection point where both sources reside in a consistent schema and governance model, enabling seamless joins and consistent access controls across Databricks-derived data and Fabric data. Other artifacts don’t serve this integration role as effectively. A data lake is raw storage without the integrated governance and semantic layer of a lakehouse. A dataflow handles transformations but isn’t the container for multi-source, governed storage. A dataset is a consumption object tied to a specific source, not the centralized integration point for combining multiple data sources.

The central idea is that a lakehouse provides the unified, governed layer you need to bring together multiple data sources. By using data pipelines to bring in Databricks data and shortcuts to reference the existing Fabric lakehouse data, you can create a single lakehouse that serves as the integrated foundation for querying and analysis. The lakehouse acts as the collection point where both sources reside in a consistent schema and governance model, enabling seamless joins and consistent access controls across Databricks-derived data and Fabric data.

Other artifacts don’t serve this integration role as effectively. A data lake is raw storage without the integrated governance and semantic layer of a lakehouse. A dataflow handles transformations but isn’t the container for multi-source, governed storage. A dataset is a consumption object tied to a specific source, not the centralized integration point for combining multiple data sources.