Skip to main content

Workspace + Genie Code

You'll learn how to build a pipeline interactively using the Databricks Workspace and Genie Code in ~10 min.

Prereqs: Build the first pipeline

Why this matters​

The fastest way to go from raw data to a working pipeline is the Databricks Workspace combined with Genie Code. Genie Code generates pipeline code from natural-language prompts, and the Workspace lets you validate, edit, and run it without leaving the browser. This is the recommended starting point before moving to code-first approaches like DABs.

How it works​

The Databricks workspace includes a pipeline editor where you can create Spark Declarative Pipelines visually. Genie Code assists by generating transformation code, suggesting data quality expectations, and auto-completing schema definitions based on your data.

Getting started​

This video walks through creating a pipeline from scratch using the Workspace and Genie Code:

Take Genie Code to the next level​

Once the basics are working, Genie Code can help with more complex transformations, joins, and data quality rules:

When to use / when not to​

Use the Workspace + Genie Code when:

  • You are building your first pipeline and want fast feedback.
  • You want to prototype a transformation before committing to code.
  • You prefer a visual, interactive workflow.

Use DABs instead when:

  • You need version-controlled, repeatable deployments.
  • The pipeline will run in staging and production environments.
  • You want to define infrastructure (schemas, volumes, warehouses) alongside the pipeline.

Next​