← Uber Interview Insights

Uber·Machine Learning Engineer·Technical Phone Screen·Senior

Senior
Jun 2026

Summary

Uber MLE interview with a focused experimentation question. Pretty technical but not in a coding sense, more like they wanted to see if you actually understand when and why you'd pick a given statistical test.

Questions Asked (2)

Q1

You're running a user-level A/B test on a continuous metric like order value or session length. Why is a two-sample t-test appropriate here, and walk through how you'd actually run it end to end.

A/B Testing & ExperimentationTechnical Trade-offs
Author's notes

I knew the mechanics well enough but stumbled a bit explaining the 'why' before the 'how'.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining why the two-sample t-test is appropriate for comparing means of a continuous metric between two independent groups, then walk through the end-to-end process from defining hypotheses to interpreting results. Emphasize practical considerations like assumption checks, sample size, and potential pitfalls in an A/B testing context.

Pro tip: Mention that while the t-test is robust to moderate violations of normality with large samples (thanks to the CLT), you should still check for extreme skew or outliers and consider transformations or non-parametric alternatives if needed. Also, highlight the importance of pre-registering the analysis plan to avoid p-hacking.

1. Justify the t-test

Explain that the two-sample t-test compares means of a continuous metric between two independent groups (control and treatment), assuming approximately normal sampling distributions or large samples. It's appropriate because we're interested in the difference in average order value or session length.

2. Define hypotheses and metrics

State the null hypothesis (no difference in means) and alternative hypothesis (two-sided or one-sided). Clearly define the metric and the unit of randomization (e.g., user-level).

3. Check assumptions and design

Verify independence, normality (or large sample size), and equal variances (or use Welch's t-test). Determine required sample size via power analysis (typically 80% power, 5% significance). Ensure random assignment and no interference between groups.

4. Run the test and compute statistics

Calculate the t-statistic and p-value using the appropriate formula (pooled or Welch). Also compute confidence intervals for the difference in means to understand practical significance.

5. Interpret results and make decisions

Compare p-value to significance level (e.g., 0.05). If significant, assess effect size and business impact. Consider multiple testing corrections if multiple metrics. Discuss limitations and next steps.

Key Points to Mention

  • Independence of observations (user-level randomization ensures this)
  • Normality assumption and the Central Limit Theorem for large samples
  • Homogeneity of variance and when to use Welch's t-test
  • Sample size and power analysis to detect a meaningful effect
  • Two-sided vs one-sided tests and pre-registration to avoid p-hacking
  • Practical significance vs statistical significance (effect size, confidence intervals)

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Under what conditions would you abandon user-level randomization and run a switchback or geographic experiment instead?

A/B Testing & ExperimentationTechnical Trade-offsProduct Analytics & Metrics
Author's notes

This is where I actually felt more comfortable.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining why user-level randomization is the default, then identify scenarios where it fails due to interference, spillover, or operational constraints. For each scenario, describe how switchback or geographic experiments address the issue, and discuss trade-offs like bias, variance, and feasibility.

Pro tip: Emphasize that switchback designs require careful time-series analysis to account for temporal trends, and geographic experiments need to handle spatial correlation and market-level heterogeneity. Mentioning these nuances shows depth beyond textbook answers.

1. Clarify the limitations of user-level randomization

Explain that user-level randomization assumes no interference between users and stable treatment effects. When these assumptions break, results can be biased.

2. Identify conditions for switchback experiments

Switchback is suitable when interference occurs within a shared resource over time (e.g., marketplace, supply chain) or when user-level randomization is infeasible due to operational constraints.

3. Identify conditions for geographic experiments

Geographic experiments are appropriate when interference spills over across users in the same region, when treatment affects an entire market (e.g., pricing, marketing campaigns), or when user-level randomization is impossible due to platform-wide changes.

4. Discuss trade-offs and analysis considerations

Compare switchback and geographic designs in terms of bias, variance, and practical implementation. Mention the need for specialized statistical methods to account for temporal or spatial correlations.

5. Conclude with a decision framework

Summarize when to choose each design based on the nature of interference, operational constraints, and the goal of the experiment.

Key Points to Mention

  • Interference/spillover effects (e.g., network effects, shared resources)
  • Operational constraints (e.g., cannot randomize at user level due to system architecture)
  • Market-level treatments (e.g., pricing, promotions, supply changes)
  • Temporal dynamics and carryover effects in switchback designs
  • Spatial correlation and geographic heterogeneity in geo experiments
  • Statistical methods like cluster-robust standard errors, time-series models, or synthetic control

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.