Fundamentals of Data Engineering

What it is

Fundamentals of Data Engineering: Plan and Build Robust Data Systems, by Joe Reis and Matt Housley, was published by O'Reilly in 2022. Instead of teaching one tool, it describes the data engineering lifecycle (generation, storage, ingestion, transformation and serving) and the undercurrents that run through all of it: security, data management, DataOps, data architecture, orchestration and software engineering.

Who it is for

Students and early-career engineers who want the mental model before the tools, software engineers moving into data, and analysts who keep hearing words like lakehouse, CDC and orchestration without seeing how they connect. It is also popular with managers who hire data engineers.

What you get from it

• A vocabulary for the whole field, so job descriptions and architecture diagrams make sense.
• A way to evaluate tools by what stage of the lifecycle they serve and what trade-offs they make, which ages far better than a tutorial for last year's product.
• Chapters on choosing technologies, on cost, and on the difference between building and buying, which are rarely covered in courses.

What it is not

It has very little code. You will not learn Python, SQL, Spark or Airflow from it. Pair it with something hands-on, such as the free Data Engineering Zoomcamp, and read the relevant chapter before each module.

Is it worth your time?

Yes, for most people starting out. The authors deliberately wrote about principles rather than products, and the book has aged well since 2022: some named tools have changed, but the lifecycle, the trade-offs and the advice about simplicity still hold. If you already work as a data engineer, skim the table of contents; you may only need a few chapters.

Free and paid options

The book is paid. You can buy it as print or ebook, or read it with an O'Reilly subscription, which many universities and employers provide. Check your library first. Prices vary by country and seller, so check the current price where you buy. Avoid unofficial PDF copies: they are pirated and often carry malware.

Time needed

About 450 pages. Reading one chapter a week alongside a practical course is a sensible pace; a focused reader can finish it in three or four weeks.

Common questions

Is there a newer edition? As of 2026-09 the 2022 first edition is current. Joe Reis has written that the book was designed to focus on timeless principles rather than tools.

Should I read it before or after Designing Data-Intensive Applications? Before. This book gives you the map of the job; Designing Data-Intensive Applications goes much deeper into how databases and distributed systems work.