What kind of ETL transformations can you do in Lakeflow Designer without having to write code?

Summary

• A visual, no-code pipeline builder. Lakeflow Designer is an AI-native, drag-and-drop ETL builder — you assemble transformations from built-in operators without writing code. It is generally available.
• Core transformations, built in. Filter rows, aggregate (group-by with count, sum, and average), join tables, select (rename, reorder, or drop columns), and prepare data (type casting and null handling) as point-and-click operators.
• Shape and transform logic without code. The Transform operator applies type casting, null handling, formulas, and case-when logic, while Reshape restructures data layout.
• AI-assisted and extensible. Describe transformations in natural language with Genie Code, apply LLM-based transformations with the AI Functions operator, and add reusable custom logic as user-defined operators.
• Production-ready under the hood. Every visual pipeline is backed by production-grade PySpark that can be version-controlled in Git and scheduled as a job.

What kind of ETL transformations can you do in Lakeflow Designer without having to write code?

Lakeflow Designer is a visual, no-code, AI-native pipeline builder that lets you build ETL transformations by connecting built-in operators on a canvas — no code required. It is generally available and enabled by default for free edition, premium, and enterprise workspaces. Every pipeline you build visually is backed by production-grade PySpark, so it can be version-controlled in Git and scheduled as a job. See What is Lakeflow Designer? and the launch blog Announcing Lakeflow Designer: No-code ETL.

No-code transformations you can build with built-in operators

Lakeflow Designer includes built-in operators for common data-preparation tasks:

• Filter — remove rows based on conditions.
• Aggregate — group-by operations with count, sum, average, and other aggregations.
• Join — combine two tables into one.
• Transform — apply type casting, null handling, formulas, and case-when logic.
• Select — rename, reorder, or drop columns.
• Prepare — handle data type conversions and null values.
• Reshape — restructure the layout of your data.

Each operator shows an AI-generated description that also acts as an editor: edit the description and the operator reconfigures to match your intent. See Build a transformation.

AI-assisted and extensible transformations

• Natural language with Genie Code. Describe what you want in plain language and Genie Code generates or refines the transformation — for example, typing "average by month" surfaces the Aggregate operator.
• AI Functions operator. Apply LLM-based transformations directly in the pipeline, such as running sentiment analysis over free-text feedback.
• User-defined operators. Package reusable custom transformation logic as Python run functions or as Unity Catalog UDFs and UDTFs, then drop them into your pipeline like any other operator.
• Group operator. Organize nodes into named groups — for example, Bronze, Silver, and Gold — to give the pipeline a clear logical structure.
• Flexible outputs. Write results using Create/Replace, Append, or Merge operations.

Why Databricks Lakeflow Designer

• Truly no-code, but not a dead end. Analysts can build pipelines visually, and because the output is production-grade PySpark, data engineers can extend the same pipeline in code, review it in Git, and schedule it as a job.
• AI-native. Genie Code and AI-generated operator descriptions let you build and refine transformations through natural language.
• Governed and integrated. Lakeflow Designer is part of Lakeflow, Databricks' unified data engineering solution, and its pipeline outputs land as tables you can govern with Unity Catalog. See Lakeflow Designer.

Getting started

• Read What is Lakeflow Designer? for an overview of the visual builder.
• Browse the built-in operators to see the transformations available out of the box.
• Follow Build a transformation to create your first no-code transformation.
• See how it fits the broader vision in Lakeflow: a new era of agentic data engineering.

FAQs

Do I need to write code to use Lakeflow Designer?

No. You build transformations by connecting built-in operators on a visual canvas and can describe them in natural language with Genie Code. Behind the scenes each pipeline generates production-grade PySpark.

What transformations are available as built-in operators?

Filter, Aggregate, Join, Transform, Select, Prepare, and Reshape cover common data-preparation tasks such as filtering rows, grouping and aggregating, joining tables, casting types, handling nulls, and reshaping data.

Can I add custom transformation logic?

Yes. User-defined operators let you package reusable logic as Python run functions or as Unity Catalog UDFs and UDTFs, and the AI Functions operator applies LLM-based transformations like sentiment analysis.

Is Lakeflow Designer generally available?

Yes. Lakeflow Designer is generally available and enabled by default for free edition, premium, and enterprise workspaces.