| title | Hands-On Lab Pack | |||
|---|---|---|---|---|
| type | labs-index | |||
| tags |
|
|||
| status | published |
Runnable, scenario-driven labs that exercise the core Databricks features tested across the certification line-up. Each lab is a single file with full code, step-by-step prose, and verification commands. Read them in order or jump to whichever feature you need to drill.
Important
Each lab assumes a Databricks workspace with Unity Catalog enabled. Pick a catalog you can write to (the examples use main — substitute your own catalog). All paths use UC volumes (/Volumes/...), not DBFS mounts.
| # | Lab | Topics exercised | Certs |
|---|---|---|---|
| 01 | Medallion ingestion | Bronze / Silver / Gold layering, Delta operations, MERGE, OPTIMIZE | DE Associate · DE Professional |
| 02 | Unity Catalog setup | Catalog / schema / volume creation, GRANT, row filters, column masks | DE Associate · DE Professional · Data Analyst · ML Associate · GenAI |
| 03 | Lakeflow Declarative Pipelines | @dlt.table, expectations, APPLY CHANGES INTO, event log |
DE Professional |
| 04 | MLflow tracking and Model Registry in UC | MLflow autologging, Model Registry in UC, version promotion | ML Associate · ML Professional |
| 05 | Mosaic AI Vector Search + RAG demo | Embeddings, Vector Search index, ResponsesAgent compound app, deployment |
GenAI Engineer Associate |
- Databricks workspace with Unity Catalog enabled and a serverless or Pro SQL Warehouse available
- A catalog you can write to (the examples use
main) - A user identity with at least
USE CATALOG+CREATE SCHEMAon the target catalog - For Lab 04: a workspace with Mosaic AI Model Serving available
- For Lab 05: Mosaic AI Vector Search and Foundation Model APIs enabled
Each lab cross-links into the relevant topic folders so you can drill deeper after running the code. The labs are exercises; the topic folders are the reference material.
The code blocks are designed to run in a Databricks notebook (PySpark / SQL / Python cells). Copy each block into a separate cell and run top-to-bottom. Verification SQL queries at the end of each section help you confirm each step worked.
Each lab ends with a Cleanup section that drops the resources it created so you don't accumulate Delta tables and Vector Search indexes across runs.