DuckDB course peels back the kernel — slides and labs posted online

Torsten Grust of the University of Tübingen has released a full set of lecture materials for a 15‑week undergraduate course titled "Design and Implementation of Database System Internals" (DiDi). It’s focused on the internals of DuckDB, the embeddable analytical database that’s been getting a lot of attention lately. The slides and auxiliary material are available in a GitHub repository, so anyone curious about what makes a modern columnar engine tick can peek under the hood.

What’s included

The repository hosts slides and supporting material that follow a 15‑week syllabus; as of March 2026, the chapter layout is published alongside the notes. You’ll need basic SQL skills to follow the red thread — most examples stick to core SELECT–FROM–WHERE–GROUP BY–HAVING patterns — and there’s a pointer to a companion course, Tabular Database Systems, for those who need an intro to the tabular model. Concise, practical, and classroom-tested: this isn’t just theory on steroids. It’s courseware you can run through.

Why it matters

Why should engineers and students care? Because educational resources that dive into real engine internals are rare, and DuckDB’s approachable codebase makes it a great teaching vehicle. Want to understand query execution, storage layout, or vectorized processing in a system people actually use? This material bridges textbook concepts and production realities. It’s a small victory for open education — and a useful stop on the roadmap for anyone building analytics systems or tooling that interfaces with embedded databases.

Find the materials on the DuckDB site and the linked GitHub repository if you want to download the slides, follow the chapters, or adapt the labs for your own class. Dig in — curiosity rewarded.

Sources: duckdb.org, Hacker News