On this page

Get Started

Learn Lakehouse Plumber the way you’ll use it: with a real project in front of you. This course scaffolds the sample project LHP ships — a complete medallion pipeline (bronze → silver → gold) over the samples.tpch dataset in every Unity Catalog workspace — then walks you through reading it, configuring it, generating its code, and deploying it to Databricks.

Prefer an agent-guided course? Copy the learning prompt into your coding agent. The separate lhp-learn skill begins with Lesson 0, explaining LHP’s model and capabilities before opening the sample files. You make the exercise edits, the agent waits after each checkpoint, and progress stays with your project. The course and its advanced topic lessons also work without website access.

The full course

You’ll work through:

  1. Install and scaffold — get the lhp command and scaffold the sample project with lhp init --sample.

  2. Configure the project — tour the four config files (lhp.yaml, substitutions, the pipeline config, databricks.yml) and set the two values the sample needs.

  3. Explore the pipelines — read the flowgroups: cloudfiles and delta ingestion, a four-transform silver chain, CDC and snapshot-CDC dimensions, a gold materialized view.

  4. Generate and inspect — run lhp generate and read the Lakeflow Python it produces.

  5. Deploy and run — ship the bundle to a workspace, run the job, and re-run it to watch incremental ingestion and CDC advance.

  6. Explore LHP Web — open the same project in the local web IDE: the pipeline as a graph, actions as forms, generation live in the browser.

  7. Where to next — start your own project, and find the features the sample didn’t cover.

The examples use checked-in sample-project fixtures. A documentation build renders those files; it does not regenerate every fixture. Command timings are illustrative and will vary between runs.