← Meta Interview Insights

Meta·Machine Learning Engineer·Onsite - System Design / Architecture·Senior

Senior
May 2026

Summary

Meta ML engineer interview focused entirely on designing a place recommendation system for a maps product. One long open-ended question that branched into a dozen sub-topics. Dense and a bit exhausting, but not unfair.

Questions Asked (7)

Q1

Design a machine learning system that recommends places a user might want to visit, like restaurants, parks, or local attractions, for a maps or local discovery product.

System DesignTechnical Trade-offsProduct Analytics & Metrics
Author's notes

This is one of those questions that feels manageable until you realize how many directions it can go.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the product context and requirements, then outline a high-level system architecture covering data sources, candidate generation, ranking, and serving. Dive into key ML components like feature engineering, model choices, and evaluation metrics, while discussing trade-offs and scalability.

Pro tip: Emphasize the importance of user context and real-time signals, and discuss how you would handle cold-start and diversity to avoid over-recommending popular places.

1. Clarify Requirements

Ask questions to understand the product goals, user base, scale, and constraints (e.g., latency, privacy). Define success metrics like CTR, dwell time, or user satisfaction.

2. Data and Features

Identify data sources: user interactions (check-ins, searches, reviews), place attributes (category, location, popularity), and contextual signals (time, weather, user location). Discuss feature engineering for user, place, and context.

3. Model Architecture

Propose a two-stage system: candidate generation (e.g., collaborative filtering, content-based, or geo-based retrieval) followed by ranking (e.g., gradient boosted trees or deep neural networks). Mention handling of cold-start and exploration.

4. Training and Evaluation

Describe offline training with historical data, using metrics like recall@k for retrieval and NDCG for ranking. Plan online evaluation via A/B tests measuring business metrics.

5. Serving and Scalability

Outline a scalable serving infrastructure with low-latency requirements, possibly using precomputed embeddings and approximate nearest neighbor search. Discuss monitoring and feedback loops.

Key Points to Mention

  • Two-stage architecture: candidate generation and ranking
  • Use of both collaborative filtering and content-based features
  • Incorporation of contextual signals (location, time, weather)
  • Handling cold-start for new users and places
  • Evaluation metrics: offline (recall@k, NDCG) and online (CTR, dwell time)
  • Scalability and latency considerations in serving

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

Q2

How would you define training labels and the prediction target for this recommendation system?

Data ModelingTechnical Trade-offs
Author's notes

Tricky because there's no single obvious label.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the recommendation scenario (e.g., feed ranking, friend suggestions) and the business objective. Then define the prediction target as the specific user action you want to predict (e.g., click, like, share) and the training labels as the observed outcomes for that action, considering time windows and negative sampling. Emphasize how these choices impact model performance and align with Meta's goals.

Pro tip: Discuss how you would handle delayed feedback and position bias, and mention that you'd validate the label definition through offline metrics and online A/B tests to ensure it drives the desired business outcome.

1. Clarify the recommendation context

Ask or state the specific recommendation surface (e.g., News Feed, Groups, Marketplace) and the primary business objective (e.g., engagement, revenue, user satisfaction).

2. Define the prediction target

Specify the exact user action to predict (e.g., click, like, comment, share, dwell time) and whether it's binary, multi-class, or regression. Consider multiple targets and how to combine them.

3. Define training labels

Describe how to derive labels from logged data: positive labels from observed actions, negative labels from non-actions, and how to handle missing data, time windows, and sampling.

4. Address challenges and trade-offs

Discuss issues like class imbalance, delayed feedback, position bias, and how label choice affects model bias and evaluation metrics.

5. Validate and iterate

Explain how you would validate the label definition offline (e.g., through holdout sets) and online (A/B tests), and iterate based on performance.

Key Points to Mention

  • Choice of prediction target (e.g., click vs. like vs. share) and its alignment with business goals
  • Label generation from implicit feedback (e.g., clicks as positives, impressions without clicks as negatives)
  • Handling of delayed feedback and attribution windows
  • Negative sampling strategies to address class imbalance
  • Position bias and how to correct for it (e.g., using inverse propensity scoring)
  • Evaluation metrics (e.g., AUC, log loss, NDCG) and online metrics (e.g., CTR, engagement rate)

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

Q3

What data sources and features would you use, and how would you engineer them for this system?

Data ModelingSystem Design
Author's notes

Went through user history, location, time of day, search queries, map interactions, place metadata.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the system's objective and constraints, then propose a mix of batch and real-time data sources, and finally detail feature engineering techniques including transformations, aggregations, and embeddings. Emphasize how features are computed, stored, and served to ensure consistency and low latency.

Pro tip: Highlight the importance of feature versioning and monitoring to prevent training-serving skew, and mention how you would leverage Meta's existing feature store infrastructure to accelerate development and ensure consistency.

1. Clarify System Goals and Constraints

Ask questions to understand the problem domain, prediction task, latency requirements, and scale. This will guide data source and feature choices.

2. Identify Data Sources

List relevant internal and external data sources, such as user interactions, content metadata, social graph, and contextual signals. Consider batch and streaming sources.

3. Design Feature Engineering Pipeline

Describe transformations for each data type: numerical (scaling, binning), categorical (one-hot, hashing, embeddings), text (TF-IDF, embeddings), and graph (node embeddings). Include temporal aggregations and windowed statistics.

4. Implement Feature Storage and Serving

Explain how features are stored (e.g., feature store) and served online (low-latency) and offline (for training). Discuss consistency between training and serving.

5. Monitor and Iterate

Outline monitoring for feature drift, data quality, and model performance. Describe how to iterate on features based on feedback.

Key Points to Mention

  • Use of Meta's feature store (e.g., FBLearner Feature Store) for consistency and reuse
  • Handling categorical features with hashing and embeddings to manage high cardinality
  • Temporal features and windowed aggregations (e.g., user's last N interactions)
  • Graph-based features from social connections (e.g., friend embeddings, interaction graphs)
  • Real-time feature computation using streaming (e.g., Kafka, Flink) for low-latency predictions
  • Feature versioning and monitoring to detect training-serving skew and drift

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

Q4

Walk me through how you'd handle cold start for both new users and new places.

Technical Trade-offsAdaptability & Ambiguity
Author's notes

New users I had covered: onboarding signals, demographic priors, popular local picks.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the two cold start scenarios (new users and new places) and their distinct challenges, then propose a unified framework that leverages side information and meta-learning. Emphasize practical trade-offs at Meta's scale, such as using content-based features for new places and few-shot learning for new users, and discuss how to evaluate and iterate.

Pro tip: Show awareness of Meta's production constraints by mentioning how you'd handle cold start in a multi-task, multi-modal setting and how you'd measure success with online metrics like CTR and engagement, not just offline AUC.

1. Clarify the problem and constraints

Ask clarifying questions to understand what 'new' means (e.g., zero interactions vs. limited), what data is available (user demographics, place attributes), and the business impact (e.g., user retention, place discovery).

2. Leverage side information and content features

For new places, use content-based features (location, category, reviews) to compute embeddings; for new users, use demographic and contextual signals to bootstrap preferences.

3. Apply meta-learning and transfer learning

Use meta-learning (e.g., MAML) to quickly adapt to new users/places from few interactions, and transfer knowledge from similar users/places via embeddings or graph neural networks.

4. Design exploration and exploitation strategy

Implement bandit algorithms (e.g., Thompson sampling) to balance showing new places to users and collecting feedback, while personalizing for new users via contextual bandits.

5. Evaluate and iterate with online metrics

Define offline proxies (e.g., recall@k for new places) and online A/B tests measuring CTR, engagement, and retention; monitor for bias and feedback loops.

Key Points to Mention

  • Content-based filtering for new places using attributes like location, category, and text descriptions.
  • Meta-learning and few-shot learning for new users to adapt from minimal interactions.
  • Hybrid approach combining collaborative filtering with side information.
  • Exploration-exploitation trade-off via multi-armed bandits for new places.
  • Evaluation metrics: offline (AUC, recall) and online (CTR, engagement, retention).
  • Scalability and production constraints at Meta (e.g., real-time inference, large-scale embeddings).

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

Q5

How would you address risks like popularity bias, feedback loops, spam, and fairness across different types of places and regions?

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

This came near the end and I was running low on steam.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around a systematic risk mitigation framework: first identify and measure each risk, then apply targeted technical and product interventions, and finally validate through rigorous experimentation with fairness-aware metrics. Emphasize that these risks are interconnected and require continuous monitoring and cross-functional collaboration.

Pro tip: Show maturity by acknowledging trade-offs—e.g., aggressive debiasing can hurt relevance—and propose a balanced approach with guardrail metrics in A/B tests. Mention that fairness definitions vary by region and require local context, demonstrating global product awareness.

1. Diagnose and Measure Risks

Define metrics for popularity bias (e.g., Gini coefficient of item exposure), feedback loops (e.g., diversity over time), spam (e.g., precision/recall of spam classifiers), and fairness (e.g., demographic parity across regions). Establish baselines and monitoring dashboards.

2. Apply Technical Mitigations

For popularity bias, use re-ranking or inverse propensity weighting; for feedback loops, inject exploration or use causal debiasing; for spam, deploy adversarial training and user reporting; for fairness, use constrained optimization or post-processing to meet region-specific fairness criteria.

3. Incorporate Product and Policy Interventions

Work with product teams to design features that promote diverse content (e.g., 'explore' tabs), implement rate limiting and verification to combat spam, and adapt policies to local norms and regulations (e.g., different fairness definitions across regions).

4. Validate with Experiments and Guardrails

Run A/B tests with guardrail metrics (e.g., fairness, spam rate, diversity) alongside engagement metrics. Use long-term holdouts to detect feedback loops and monitor regional disparities. Iterate based on results.

5. Monitor and Iterate Continuously

Deploy real-time monitoring for drift and anomalies, set up alerts for fairness violations, and establish a feedback loop with cross-functional teams to update models and policies as new risks emerge.

Key Points to Mention

  • Popularity bias: use debiasing techniques like inverse propensity scoring or diversity-aware ranking.
  • Feedback loops: introduce exploration (e.g., epsilon-greedy) and use causal inference to break loops.
  • Spam: combine ML classifiers with user reports and adversarial training; consider adversarial attacks.
  • Fairness: define metrics per region (e.g., equal opportunity) and use constrained optimization; be aware of legal and cultural differences.
  • A/B testing: include guardrail metrics for fairness, spam, and diversity; use long-term holdouts.
  • Cross-functional collaboration: partner with policy, product, and regional teams to tailor solutions.

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

Q6

How would you set up online serving for this system given real-time context like current location and time of day, and what latency constraints would you design around?

System DesignTechnical Trade-offs
Author's notes

Talked about pre-computing candidate sets and doing lightweight real-time re-ranking with context features.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the system's requirements and scale, then propose a two-stage serving architecture with a fast retrieval layer and a more complex ranking layer. Emphasize how real-time features like location and time are fetched and incorporated with low latency, and discuss trade-offs between latency and model complexity.

Pro tip: Quantify latency budgets for each stage (e.g., retrieval <50ms, ranking <100ms) and mention techniques like caching, approximate nearest neighbor search, and feature pre-computation to meet those budgets.

1. Clarify Requirements and Scale

Ask about the system's purpose, expected QPS, user base, and latency SLOs. Understand what real-time context is available and how critical it is for predictions.

2. Design Two-Stage Serving Architecture

Propose a retrieval stage to quickly narrow down candidates (e.g., using embeddings and ANN) and a ranking stage to score them with a more complex model. Explain how real-time features are integrated in each stage.

3. Handle Real-Time Features

Describe how to fetch and process location and time-of-day features with low latency, using in-memory stores, precomputed aggregates, or streaming pipelines. Discuss fallbacks for missing data.

4. Define Latency Constraints and Optimizations

Set latency budgets for each component (e.g., feature fetch <10ms, retrieval <50ms, ranking <100ms) and propose optimizations like caching, model quantization, and parallel processing.

5. Discuss Trade-offs and Monitoring

Acknowledge trade-offs between latency, accuracy, and cost. Mention the need for monitoring and A/B testing to ensure the system meets business goals.

Key Points to Mention

  • Two-stage architecture: retrieval and ranking
  • Real-time feature serving with low-latency stores (e.g., Redis, feature stores)
  • Approximate nearest neighbor (ANN) for fast retrieval
  • Latency budgets and SLOs (e.g., p99 < 200ms)
  • Caching and precomputation of features
  • Trade-offs between model complexity and latency

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

Q7

How would you run experiments and manage feedback loops in a live recommendation system like this?

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

A/B testing with holdout groups, measuring both engagement and diversity metrics, watching for cannibalization between recommendation surfaces.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by outlining a structured experimentation framework that covers offline evaluation, online A/B testing, and continuous feedback integration. Emphasize the importance of defining clear metrics, ensuring statistical rigor, and iterating quickly based on results. Highlight how you would handle challenges specific to live recommendation systems, such as feedback loops and cold start.

Pro tip: Demonstrate awareness of the trade-off between exploration and exploitation, and mention techniques like interleaving or multi-armed bandits to efficiently test multiple models. Also, stress the importance of guardrail metrics to catch regressions in user experience.

1. Define Objectives and Metrics

Clearly articulate the goal of the experiment (e.g., increase click-through rate or user engagement) and select primary and secondary metrics. Include guardrail metrics to monitor for negative side effects.

2. Offline Evaluation and Simulation

Use historical data to evaluate candidate models offline, employing techniques like counterfactual evaluation or replay to estimate performance before live testing. This helps filter out poor models and reduces risk.

3. Online A/B Testing

Design and run controlled online experiments, randomizing users into control and treatment groups. Ensure proper sample size, duration, and statistical power to detect meaningful differences.

4. Monitor and Analyze Results

Continuously monitor experiment metrics for statistical significance and practical impact. Analyze segment-level performance and investigate any unexpected effects or interactions.

5. Iterate and Integrate Feedback

Based on results, decide whether to launch, iterate, or abandon the change. Incorporate learnings into the next cycle, and consider using bandit algorithms for continuous optimization.

Key Points to Mention

  • A/B testing best practices: randomization, control groups, statistical power, and avoiding peeking
  • Offline evaluation methods: counterfactual evaluation, replay, and simulation
  • Exploration-exploitation trade-off and techniques like multi-armed bandits or interleaving
  • Guardrail metrics to detect regressions in user experience or system health
  • Handling feedback loops and biases in live recommendation systems
  • Scalability and automation of experimentation pipelines for rapid iteration

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