Question bank

Interview questions, with guidance on how to answer

834 questions across 64 roles. Open one to see what a strong answer includes and common slips, then drill it out loud with Ava. New to answer structure? Start with the learning hub.

834 questions.

TechnicalMid levelData Engineer
How do you structure dependencies and retries in an orchestration tool like Airflow or Dagster so that one failure doesn't cascade into a mess?
Tests: Pipeline design & orchestration
TechnicalMid levelData Engineer
Walk me through how you would choose between a cloud warehouse like Snowflake or BigQuery and a lakehouse setup for a new analytics platform.
Tests: Cloud data platforms
BehaviouralMid levelData Engineer
Tell me about a time an analyst or business team asked for data in a way you thought was the wrong solution. How did you handle it?
Tests: Stakeholder communication
TechnicalSeniorData Engineer
Imagine our warehouse bill doubled in six months. How would you investigate and bring it down without breaking what people depend on?
Tests: Performance & cost optimisation
TechnicalSeniorData Engineer
How do you approach designing a data model that several teams will share, when each team defines a key metric like active customer differently?
Tests: SQL & data modelling
BehaviouralSeniorData Engineer
Tell me about the most serious data incident you've owned. What happened, how did you communicate it, and what changed afterwards?
Tests: Ownership & incident response
TechnicalAny levelData Engineer
How do you deal with upstream application teams changing their database schema and breaking your pipelines?
Tests: Data quality & testing
BehaviouralAny levelData Engineer
When you inherit a pipeline with no documentation, what do you do in your first couple of weeks with it?
Tests: Ownership & incident response
TechnicalEntry levelData Scientist
How would you explain a p-value to a product manager who has never taken a statistics course?
Tests: Statistics & experimentation
TechnicalEntry levelData Scientist
You get a dataset where about a fifth of the values in an important column are missing. How do you decide what to do?
Tests: Data wrangling & SQL
TechnicalEntry levelData Scientist
What is overfitting, and how would you spot it in a model you just trained?
Tests: Predictive modelling
TechnicalMid levelData Scientist
Tell me about a split test or experiment you designed or analysed. How did you choose the success metric and the sample size?
Tests: Statistics & experimentation
TechnicalMid levelData Scientist
You're building a fraud model where only one in a thousand transactions is fraud. How would you evaluate whether it's any good?
Tests: Model evaluation
TechnicalMid levelData Scientist
A stakeholder asks you to build a churn model. What questions do you ask before writing any code?
Tests: Problem framing
BehaviouralMid levelData Scientist
Tell me about a time your analysis showed something leadership didn't want to hear. How did you present it?
Tests: Communicating findings
CaseSeniorData Scientist
When you can't run an experiment, how do you estimate whether a new feature actually caused an increase in retention?
Tests: Causal reasoning
TechnicalSeniorData Scientist
How do you decide between a simple interpretable model and a more complex one when the complex one is only slightly more accurate?
Tests: Predictive modelling
BehaviouralSeniorData Scientist
Tell me about a time you had to get engineering and product teams to change direction based on data. How did you build agreement?
Tests: Collaboration & influence
TechnicalAny levelData Scientist
How do you check that the data you're analysing actually means what you think it means?
Tests: Data wrangling & SQL
TechnicalAny levelData Scientist
After a model goes live, how would you know it's starting to get worse?
Tests: Model evaluation