314,552 interview questions from 6,000+ companies.
Tests influence without authority: aligning stakeholders through data, empathy, and ownership to drive a decision and measurable outcome.
Approach for maintaining data quality and integrity across ETL pipelines.
Tests communication of complex analytics to nontechnical stakeholders, with emphasis on influence, clarity, and driving action from insights.
Tests conflict resolution in technical leadership: mediating disagreement, driving a decision, and preserving team trust and execution.
Investigate why a key KPI moved the wrong way after a product change and separate signal from noise.
Tests prioritization under pressure, stakeholder management, and decision-making when multiple teams compete for limited analyst capacity.
Explain the bias-variance tradeoff and how it guides model choice, regularization, and generalization performance.
Tests adaptability under changing requirements, with emphasis on prioritization, ambiguity management, and ownership during a technical pivot.
Choose hyperparameters with cross-validation and validation metrics, while balancing bias, variance, and overfitting.
Reason about sample size, power, and minimum detectable effect before launching an experiment.
Choose a decision threshold for a classifier using precision, recall, calibration, and confusion matrix tradeoffs.
Explain how to evaluate a regression model with RMSE and MAE, and how to interpret the tradeoff between average and large errors.
Explain precision vs recall and when business context should push you to optimize one metric over the other.
Explain what a confidence interval means and how to communicate it to a non-technical stakeholder.
Framework for deciding whether to ship a feature as an MVP or wait for a more complete launch.
How to tell whether a model's predicted probabilities are well calibrated, and what the business impact is.
Use LAG to compare an operational KPI week over week and surface absolute and percentage change.
Use a Splice A/B test to explain p-value and statistical power in plain English, then quantify both with a two-proportion test.
Use lift values at different targeting depths to choose between a model-based and rule-based campaign strategy under a 20% budget cap.
Reason about experiment design when treatment can spill over across users and bias user-level randomization.
