Databricks is bringing a visual, no-code approach to data preparation with the public preview of Lakeflow Designer. This new offering aims to democratize data access by allowing analysts and domain experts, not just data engineers, to prepare and analyze data through a drag-and-drop interface and natural language prompts.
The platform, built directly on Databricks and governed by Unity Catalog data governance, keeps data in place while providing essential features like lineage and permissions from the outset. This integration promises a more streamlined and secure data workflow.
Lakeflow Designer leverages Databricks' Genie Code AI data transformation capabilities. Users can describe desired transformations in plain English, with the AI generating or modifying workflows. Unlike external tools, Genie Code can access rich metadata from Unity Catalog, ensuring more context-aware AI suggestions.
Visualizing AI-driven transformations
Each step in Lakeflow Designer is represented as a visual operator, offering users a clear, step-by-step preview of how data changes. This transparency aims to make AI-generated transformations easier to review and trust.
The tool generates production-ready Python code behind every visual transformation. This code can be reviewed, version-controlled, and integrated into existing production pipelines, reducing the need for IT to rebuild work done in separate systems.