Databricks Training

A self-paced course in data engineering and data architecture — read it, or listen on the go. Built from Databricks' own documentation, release notes, and exam guides.

Plays straight through all 51 lessons, in order, across both tracks. Keeps going while your screen is off.

Data Engineering

E0Platform Fundamentals & Architecture Review

The mental model everything else hangs off — 5 lessons · 38 min audio · 6% of exam

E1Data Ingestion and Loading

Getting data in, exactly once, without babysitting it — 5 lessons · 48 min audio · 21% of exam

E2Data Transformation and Modeling

Turning raw rows into tables people can trust — and query fast — 5 lessons · 75 min audio · 22% of exam

E3Working with Lakeflow Jobs

Orchestration that survives contact with production — 5 lessons · 78 min audio · 16% of exam

E4Governance and Security

Who can see which rows, who granted it, and how you prove it — 5 lessons · 70 min audio · 15% of exam

E5Implementing CI/CD

Stop clicking things — describe the workspace as code and deploy it from a pipeline — 5 lessons · 51 min audio · 10% of exam

E6Troubleshooting, Monitoring, and Optimization

Find the actual bottleneck before you touch anything — 5 lessons · 66 min audio · 10% of exam

E7Apache Spark Internals and Optimization

What the engine is actually doing, and why that makes it slow — 5 lessons · 47 min audio

Data Architecture

A0Lakehouse Architecture Foundations

The argument, the mechanism, the tradeoffs — and how to defend them in a design review — 5 lessons · 71 min audio

Reference

R0Databricks Services & Terms: The Complete Glossary

Every product name and every term that shows up in a meeting, defined once, correctly — 6 lessons · 69 min audio

How to use this

🎧 Commuting

Open a module and hit Play. Narration is written for listening — it never says "as you can see here". Lock-screen controls work.

⌨️ At a keyboard

Read the lesson notes, then run the module lab against your own workspace.

📄 Interview tomorrow

Go straight to the cheat sheets. They print to one page.

👥 Teaching a team

Every lesson ends with facilitator notes: timings, demos, and the errors learners actually make.

Why this exists

Databricks renamed at least eight products between June 2025 and July 2026, so most training material on the internet is wrong on naming and some is wrong on code. Every fact here links to Databricks' own documentation and carries the date it was verified. Where the docs contradict themselves, this course says so instead of picking one and pretending it's settled.

51 lessons · 613 min of narration · verified 2026-07-30