Microsoft Fabric Data Engineer Associate DP-700 Practice Question

A notebook is the most suitable Fabric item for this scenario because notebooks let you write and run custom code in languages such as Python or Scala, install and use external libraries (for example, scikit-learn), call REST endpoints, and perform complex iterative logic that is common in data science and advanced data preparation. Notebooks can be scheduled directly in Microsoft Fabric and can read from or write to lakehouses, providing seamless integration with other Fabric artifacts.

Pipelines are optimized for orchestrating discrete activities and simple data-movement tasks; they cannot run arbitrary Python code with custom libraries inside a single activity without wrapping the code in another service such as a notebook. Dataflows are designed for no-code or low-code transformational logic and do not support installing custom Python packages or complex control flow. SQL stored procedures in a warehouse cannot natively execute external REST calls or Python libraries and are limited to SQL-based operations. Therefore, creating a notebook best meets all the requirements.