picked.
Log inPost a role
picked.ai/hire/data-engineer/interview-questions
30 data engineer
interview questions that actually work.
Pulled from the Neuroworx item bank: nine years of calibration against twelve-month performance outcomes on 14,083 data engineers. Sorted by stage (screen, assessment, on-site) and level (IC1 to IC5). Each question comes with what to listen for, what to ignore, and the failure mode it is designed to catch.
30
questions
4
stages
5
levels
14k
hires of validity data
Screen ↓Role-fit ↓On-site ↓Anti-pattern questions ↓
Stage 01 · Screen
Twelve minutes. Ten questions.
The screening conversation. Picked runs this with an AI voice; this is what a human screen would look like with the same rubric. Time-box hard. 60 seconds per answer.
10 questions
01
Tell me about the last pipeline you owned past launch. What broke first in production, and what did you do?
ownershipspecificity
Listen for
A specific pipeline. A real failure (late data, schema drift, duplicate keys). A fix that names a trade-off.
Ignore
Adjective-heavy answers about "reliable" or "automated" with no incident.
catches · CV inflation. Pipelines they touched but did not own.
02
Pick a model or table in your current warehouse you think is wrong. Why has it not been fixed?
judgementpolitics
Listen for
A specific table. An honest reason for the not-yet (downstream consumers, cost of backfill, fear of breaking a dashboard).
Ignore
"Everything is fine."
catches · Cannot hold an opinion without making it personal.
03
Walk me through the most recent data product you took from spec to production.
scopeagency
Listen for
The full arc: source contract, transform, contract with the analyst, monitoring. Specific dates and row counts.
Ignore
A list of tools used.
catches · Engineers who only own the dbt model and not what it feeds.
04
What is the smallest schema decision you have fought hardest for?
convictiontaste
Listen for
A real example. A primary-key choice, a nullable column, a timestamp type. They say what they would do differently.
Ignore
Big architecture flexes about lakehouses.
catches · No discrimination between large and small decisions.
05
Pick a data product you use daily (a dashboard, a feed, a report) and tell me how it probably works under the hood.
curiositysystems
Listen for
Reasonable guesses at sources, transforms, refresh cadence, caching. They name what they do not know.
Ignore
Marketing copy from a vendor site.
catches · Engineers who have never been curious about a system someone else built.
06
Last time an analyst pushed back on a model you owned. Walk me through it.
commscross-team
Listen for
A concrete disagreement. Whether they shifted, and what changed their mind.
Ignore
"I always listen to analysts." Means nothing.
catches · Engineers who treat analysts as ticket-filers.
07
What is your favourite pipeline failure you have hunted down?
reliabilitywar stories
Listen for
A specific failure with a real root cause. Their role. What changed in the design after.
Ignore
A dramatic story with no resolution.
catches · Engineers who have never been close enough to the data to have one.
08
Tell me about a transformation tool, warehouse, or orchestrator you tried and rejected.
tastecritical thinking
Listen for
A real evaluation. A specific reason to reject (cost, operational pain, team familiarity). What they used instead.
Ignore
A tool they are "exploring".
catches · CV as the boundary of what they have considered.
09
How do you onboard onto an unfamiliar warehouse?
judgementgenerality
Listen for
A sequence: read the source-to-mart lineage, find the most-queried tables, run an old failed job, ask one specific question.
Ignore
"I read all the dbt docs first." Does not scale.
catches · Engineers who freeze without a data catalogue.
10
What is one thing you want in your next role that you would not apply for a role without?
stage fitseriousness
Listen for
Something specific. A particular scale, a particular kind of data, a particular analyst team.
Ignore
"Impact." "Growth." "Ownership."
catches · Candidates who are not sure why they are looking.
Stage 02 · Role-fit assessment
A scoped task. A scored rubric.
One realistic task. We score the writeup, not the polish. The candidate has the take-home equivalent of 60 minutes.
8 questions
01
Design the pipeline for a new event source emitting 200 million rows a day. Sketch components, identify three failure modes, and name what you would build first.
systems designIC3+
Listen for
Capacity reasoning. A real first cut. Three actual failure modes (late data, schema drift, duplicate keys, partition skew).
Ignore
A perfect-looking architecture diagram.
catches · Architecture astronauts. Comprehensive on paper, helpless on call.
02
Here is a 250-line dbt model from a real warehouse. Refactor it. Three bullets on what you changed and why.
code qualityIC2+
Listen for
CTE structure, materialisation choice, tests added, columns dropped. The bullets reveal taste.
Ignore
Renaming and adding more comments than tests.
catches · Cargo-cult refactors.
03
Pick one of three small pipeline bugs we describe. Reproduce, fix, write the post-mortem. 60 minutes.
craftIC2+
Listen for
How they reach the repro. What they do not change in the model contract. The shape of the post-mortem.
Ignore
Time on environment setup.
catches · Engineers who cannot name what they do not know.
04
Read this 3-page proposal for a new mart. Write three questions for the author and one push-back.
judgementIC3+
Listen for
Questions that show they have read the doc. A push-back that engages with the contract trade-off the author made.
Ignore
Stylistic edits.
catches · Engineers who cannot engage with someone else's design.
05
Estimate the daily warehouse cost of running this pipeline. Show your working in three sentences.
cost-awareIC2+
Listen for
Reasonable orders of magnitude. They name partitioning, scan size, and refresh cadence as the drivers.
Ignore
Spreadsheets.
catches · Engineers who only know how to build, never how to estimate.
06
Take this real merge request changing a dbt model with downstream consumers. Decide if you would approve, request changes, or close. Write the review.
code reviewIC2+
Listen for
Substantive comments on the contract change, the missing test, the downstream impact.
Ignore
Line-level comments on SQL formatting only.
catches · Reviewers who cannot prioritise.
07
Write the on-call runbook for the pipeline you sketched in question 1.
operabilityIC3+
Listen for
They imagine being paged at 3am. Specific signals, specific commands, specific decisions about backfill versus skip.
Ignore
Cargo-culted runbook templates.
catches · Engineers who can build but cannot imagine operating.
08
In 200 words, why might the pipeline you sketched in question 1 be the wrong choice?
humilityIC4+
Listen for
Genuine engagement with the alternative. A real "I might have used a streaming approach" or "I should have started with one daily batch".
Ignore
A second pitch for the original design.
catches · Lack of perspective on their own choices.
Stage 03 · On-site (after Picked)
Twelve questions you will still want to ask in person.
Picked screens, scores, and shortlists. These are the questions worth asking with a human in the room: the calibration questions, the dealbreakers, the chemistry probes.
12 questions
01
Where, in the work, do you want to grow most this year?
growthmanager fit
Listen for
A specific gap. A plan, even tentative. A name of someone they would learn from.
Ignore
"I want to become a staff engineer."
catches · Engineers without a learning agenda.
02
Tell me about a time you disagreed with a manager. What happened?
authoritymanager fit
Listen for
A real disagreement. The mechanics, not the moral.
Ignore
"I have never disagreed with a manager."
catches · Engineers who cannot hold opinions in the face of authority.
03
What is the most uncomfortable feedback you have received and what did you do with it?
self-awareness
Listen for
A specific piece of feedback. The change they made. The thing they still struggle with.
Ignore
"I take feedback well."
catches · Defended self-narrative.
04
Walk me through a pipeline or project you wish had failed faster.
judgementoperating
Listen for
Honesty. A specific moment they could have called it. What stopped them.
Ignore
A pitch for the project being secretly worth doing.
catches · Sunk-cost thinkers.
05
What is a strong opinion about data systems you have recently changed?
intellectual humility
Listen for
A specific opinion. A specific reason. They name the bug, post, or person that changed their mind.
Ignore
"My mind is always open."
catches · Closed-loop thinkers.
06
Pick two senior data engineers or analysts you admire from your last role. What do they do differently?
taste
Listen for
Concrete habits. Habits they have adopted, habits they have not.
Ignore
Pure praise.
catches · Engineers without taste for peers.
07
Tell me the last technical thing you read about data outside your job.
curiosity
Listen for
A specific paper, blog, or talk. They tell you what they thought, not just that they read it.
Ignore
A textbook they "always mean to get to".
catches · Engineers who do not think outside their stack.
08
When are you most productive?
operating model
Listen for
A specific time-of-day and environment. A self-aware answer about energy.
Ignore
"I am always productive."
catches · Engineers without self-instrumentation.
09
Where would you rather be in three years?
careerretention
Listen for
A direction (deeper IC, drifting toward platform or analytics engineering, people-management) and a reason.
Ignore
"Wherever the company needs me."
catches · Drifting engineers.
10
If you join, what would you want your first week to look like?
agencyonboarding
Listen for
A specific plan. Often: shadow on-call, sit with an analyst, ship one tested model, read three post-mortems.
Ignore
"Whatever you suggest."
catches · Engineers without an onboarding instinct.
11
What would make you leave us within six months?
dealbreaker
Listen for
A specific irritant. A specific working condition. A specific kind of management.
Ignore
"As long as the work is good."
catches · Hidden dealbreakers, surfaced post-offer.
12
What is one question you want to ask our most cynical analyst?
probingcuriosity
Listen for
A real question. Usually about a quiet thing: "Which dashboard do you not trust?" or "What model do you work around?"
Ignore
A softball or a re-pitch of their interest.
catches · Candidates who do not want to know what is wrong.
The anti-pattern set
Eight questions that look smart
but tell you nothing.
"What is your biggest weakness?"
You will get a strength-shaped weakness. We have asked this 47,000 times. It catches no-one. Replace with: "What is the most uncomfortable feedback you have received?".
"Where do you see yourself in five years?"
Either a rehearsed answer or a stalled one. Both useless. Replace with: "Where would you want to be in three years?"
"Tell me about yourself."
Wastes the first three minutes on the CV they already gave you. Replace with: "Walk me through the most recent thing you shipped end-to-end."
"Why this company?"
Generates polished mission-talk. Replace with: "What about this role made you apply that would not have made you apply elsewhere?"
"Are you a team player?"
No-one says no. Replace with: "Tell me about a time a teammate disagreed with you and how you handled it."
"How do you handle stress?"
No-one says badly. Replace with: "Tell me about your last production incident and your precise role."
"How would you reverse a linked list?"
Probes nothing we care about. We removed it from the bank in 2019. Replace with: "Refactor this 200-line file and tell me what you changed and why."
"If you were an animal, which animal would you be?"
You know what we are going to say. Replace with: anything else.
Or, let us ask
We will ask these for you.
By Friday.
Picked runs the screen, the assessment, and the first-round interview against this exact item bank. You meet the three finalists in person, with these on-site questions in hand.
$0.99 per AI-vetted candidate. First 50 free.
Companion pages
picked.
The AI-native replacement for inbound hiring. For the hiring manager, not HR.
© 2026 Picked · Made in London · Three finalists in forty-eight hours.
Data engineer interview questions · Picked.ai