← Meta Interview Insights

Meta·Data Scientist·Onsite - System Design / Architecture·Senior

Senior
Apr 2026

Summary

Meta DS interview focused entirely on a machine learning system design problem around ranking shop ads in a marketplace app. Pretty open-ended, which I wasn't fully prepared for. The question sprawled across modeling, evaluation, and failure modes all at once.

Questions Asked (4)

Q1

Design an end-to-end ML system for ranking and recommending shop ads in a marketplace app. Walk through prediction targets, training labels, features, model choice, evaluation strategy, and at least three failure modes.

System DesignTechnical Trade-offsA/B Testing & Experimentation
Author's notes

This was basically the whole interview crammed into one prompt.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around the ML lifecycle: define the prediction target (e.g., click or conversion probability), then describe label generation, feature engineering, model selection, and evaluation. Emphasize trade-offs between offline metrics and online A/B tests, and proactively discuss failure modes like position bias and cold start.

Pro tip: Frame the system as a multi-stage ranking pipeline (retrieval → ranking → re-ranking) and highlight how you'd handle counterfactual logging and delayed feedback, which are common in ads.

1. Define Prediction Targets and Labels

Choose targets like CTR, CVR, or GMV, and explain how to construct training labels from logged events, addressing delayed conversions and negative sampling.

2. Design Features and Model Architecture

List user, ad, and context features, then propose a model (e.g., deep neural network with embeddings) and justify why it suits the ranking task.

3. Outline Training and Serving Pipeline

Describe data collection, feature store, training schedule, and low-latency serving with a two-stage retrieval and ranking system.

4. Evaluation Strategy

Cover offline metrics (AUC, NDCG), online A/B testing with guardrail metrics, and long-term holdout experiments.

5. Identify Failure Modes and Mitigations

Discuss at least three failure modes (e.g., position bias, feedback loops, cold start) and propose solutions like inverse propensity weighting or exploration.

Key Points to Mention

  • Position bias and how to correct it with inverse propensity scoring or unbiased learning.
  • Cold start problem for new ads and users, and solutions like content-based features or bandits.
  • Delayed feedback in conversion labels and techniques like survival analysis or importance weighting.
  • Multi-objective optimization (e.g., balancing CTR and CVR) and how to combine them in ranking.
  • Online evaluation with A/B tests, including guardrail metrics and long-term effects.
  • Scalability and latency constraints in a production ranking system.

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

Q2

How would you handle position bias in your training data, and what are the risks if you ignore it?

Data ModelingTechnical Trade-offs
Author's notes

Came up as a follow-on after I glossed over it.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining position bias and its sources in your data, then outline a structured mitigation strategy that includes detection, correction, and validation. Emphasize the risks of ignoring it, such as skewed model performance and ethical issues, and tie your answer to Meta's scale and user impact.

Pro tip: Mention that position bias often interacts with other biases (e.g., selection bias) and that mitigation should be evaluated using online metrics like CTR and fairness audits, not just offline metrics.

1. Define and Detect Position Bias

Explain what position bias is (e.g., users click top results more regardless of relevance) and how to detect it using methods like randomization, swapping, or click models.

2. Mitigation Strategies

Describe techniques to correct for position bias, such as inverse propensity scoring (IPS), position-based click models, or adding position as a feature during training but removing it at inference.

3. Validation and Monitoring

Outline how to validate the effectiveness of bias mitigation using A/B tests, counterfactual evaluation, and fairness metrics, and set up ongoing monitoring.

4. Risks of Ignoring Position Bias

Discuss the consequences: biased models that reinforce existing biases, poor generalization, reduced user trust, and potential ethical and legal issues.

5. Meta-Specific Considerations

Tie your answer to Meta's context: large-scale data, diverse user base, and the need for robust, fair AI systems that align with Meta's responsible AI principles.

Key Points to Mention

  • Definition and sources of position bias (e.g., presentation order, UI layout)
  • Detection methods: randomization, swapping, click models
  • Correction techniques: IPS, position as a feature, counterfactual learning
  • Evaluation: offline metrics vs. online A/B tests, fairness audits
  • Risks: feedback loops, echo chambers, reduced model performance, ethical concerns
  • Meta's scale and commitment to responsible AI

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

Q3

What features would you use for this ranking problem, and how do you handle cold start for new shops with no historical data?

Data ModelingProduct Analytics & Metrics
Author's notes

Cold start is such a classic problem and I still managed to give a mediocre answer.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the ranking objective and available data, then propose a mix of shop-level, user-level, and interaction features. For cold start, describe a fallback strategy using content-based features and exploration, and explain how you would transition to personalized models as data accumulates.

Pro tip: Emphasize that cold start is not just a modeling problem but also a product and data collection opportunity—suggest logging impressions and using bandit algorithms to efficiently gather feedback.

1. Clarify the problem and data

Ask about the ranking goal (e.g., relevance, conversion), the target metric, and what data is available for shops and users. This ensures your feature suggestions are aligned with business objectives.

2. Propose features for warm shops

List features such as shop historical performance (CTR, conversion rate), user-shop interaction history, contextual features (time, location), and shop attributes (category, price range).

3. Address cold start with content-based features

For new shops, rely on intrinsic shop attributes (e.g., category, description embeddings, images) and user preferences to compute similarity-based scores. Also consider using user demographics and context.

4. Implement exploration and fallback strategies

Use multi-armed bandits or epsilon-greedy to explore new shops and gather data. Fallback to popularity or editorial rankings if no personalized signal is available.

5. Plan for transition and evaluation

Describe how to gradually incorporate learned features as data accumulates, and how to evaluate cold-start performance (e.g., A/B tests, offline metrics like coverage and diversity).

Key Points to Mention

  • Feature engineering: shop-level, user-level, interaction, and contextual features
  • Cold start: content-based filtering, embeddings, and similarity
  • Exploration vs exploitation: bandit algorithms for new shops
  • Fallback mechanisms: popularity, editorial, or category-based ranking
  • Evaluation metrics: CTR, conversion, coverage, diversity, and long-term value
  • Data collection: logging impressions and feedback for new shops

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

Q4

How would you evaluate this system offline and online, and what guardrails would you put in place during an A/B test?

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

Felt more comfortable here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the system's goal and success metrics, then outline a two-phase evaluation: offline using historical data and online via A/B testing. For the A/B test, describe key guardrails like sample size, duration, and monitoring for novelty effects and metric regressions.

Pro tip: Emphasize that offline metrics are proxies and may not perfectly correlate with online impact; always validate with a small-scale online test before full rollout. Also, mention the importance of pre-registering the analysis plan to avoid p-hacking.

1. Define objectives and metrics

Clearly state the system's purpose and identify primary success metrics (e.g., CTR, conversion) and guardrail metrics (e.g., latency, user satisfaction).

2. Offline evaluation

Use historical data to simulate the system's performance. Employ techniques like cross-validation, holdout sets, and counterfactual evaluation to estimate impact.

3. Online evaluation via A/B test

Design an A/B test with proper randomization, control, and treatment groups. Determine sample size and duration based on power analysis.

4. Implement guardrails

Set up automated alerts for metric regressions, ensure no interference between groups, and monitor for novelty and primacy effects. Use sequential testing if needed.

5. Analyze and iterate

After the test, analyze results with statistical rigor, check for heterogeneous treatment effects, and decide on rollout or iteration.

Key Points to Mention

  • Offline metrics: AUC, precision/recall, RMSE, etc., depending on the system type.
  • Online metrics: CTR, conversion rate, engagement, revenue, and guardrail metrics like latency and error rates.
  • A/B test design: randomization unit, sample size calculation, power, significance level, and duration.
  • Guardrails: automated monitoring, kill switches, sequential testing, and avoiding peeking.
  • Common pitfalls: novelty effects, network effects, and Simpson's paradox.
  • Statistical methods: hypothesis testing, confidence intervals, and Bayesian approaches.

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