D8LooP · Interview preparation
Prepare for the Data Engineering interview.
It is four interviews wearing one job title. Know which one you are in, and what it is scored on.
01
The six rounds
Nearly every data engineering company runs this same loop, in this order. Only the names on it change.
02
Interview questions, by subject
1648 of them. Each one carries the answer, the answer that gets you rejected, and the follow-up they ask next.
- 160SQLquestions
- 192Pythonquestions
- 140Data Modelingquestions
- 169Data Pipelinequestions
- 150Apache Sparkquestions
- 130dbtquestions
- 132Airflowquestions
- 80Apache Kafkaquestions
- 80Lakehouse Table Formatsquestions
- 20Cloudquestions
- 80AWSquestions
- 80GCPquestions
- 80Azurequestions
- 105Data Processing Algorithmsquestions
- 50Behavioralquestions
03
50 companies, round by round
The questions candidates were actually asked — and what got people rejected.
04
Or follow a plan
18 staged plans. Every step is a chapter or a reviewed exercise that already ships — and it ticks itself off as you go.
Build Your Foundation
Interviewing
- Data Engineer · 0–2 yearsYou can prepare every round of an early-career loop and defend your own project.
- Data Engineer · 2–4 yearsYou can prepare every round of a mid-level loop: SQL, Python, modeling, design, Airflow, Spark, behavioral.
- Senior Data Engineer · 4–8 yearsYou lead a design, defend one model, and tell ownership stories with real stakes.
- Staff Data Engineer · 8+ yearsYou show org-scope impact: designs other teams depend on, and stories with numbers.
- Ace the SQL roundYou can take a vague prompt to a defended answer and explain the cost.
- Ace the Python roundYou can transform a real dataset, defend the edge cases, and be reviewed.
- Ace the Data Pipeline roundYou can design a pipeline and defend the reruns, the late data and the cost.
- Ace the Data Modelling roundYou can design a model from a vague prompt and defend every choice in it.
- Ace the Spark roundYou can narrate any Spark job from its plan, and fix it with evidence.
- Senior / staff depthYou argue cost, failure and change without being asked to.
- Interview in 3 weeksYou can pass a SQL screen, survive the Spark round, and defend a design out loud.
Hands on Concepts
- Design the warehouseYou own a model that loads nightly and that other teams trust.
- Real-time & streamingYou can run a stream in production and explain what it guarantees.
- Own the on-call pagerYou can take the pager and give the numbers back without a second incident.
- Make Spark jobs fastYou can make a slow Spark job cheaper and prove the answer did not change.
05
Then sit a live one
A timed interview with a code, across 14 modules, answered in the real studios and graded on what actually ran. Run one for a candidate, or sit one yourself.
01 · You are the interviewer
Conduct one
- PickModules, then questions: from the library, or your own.
- LockEach module, once its questions are ready.
- FinalizeA window, a code, a link to share.
02 · You are the candidate
Join one
- PasteThe code, or the whole link. A free account is enough.
- StartThe clock runs from the moment you press it. One attempt.
- AnswerSQL and Python run, Spark opens a notebook, a design goes on the canvas.
- Hire signalStrong hire · Hire · Lean no · No hire, from what ran.
- Per moduleStrong, Solid or Weak, with what stood out.
- DebriefStrengths, gaps, and what to ask next time.
- Your verdictOverrides the automatic grade on any question.
06
Try it yourself
You do not learn to design a system by reading about one. So you build it, submit it, and defend it — the way the round itself works.
- 01BuildOn the canvas. Entities, keys, grain — your decisions, not a multiple choice.
- 02SubmitOne click. No account, no card, no trial clock.
- 03ReviewScored on named axes and told what a senior engineer would have asked next.