← Uber Interview InsightsMachine Learning Engineer

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

Senior
Jun 2026

Summary

ML system design round at Uber for an MLE role. The whole session was a single deep-dive into building a destination autocomplete system, and it went pretty far into the weeds on ranking, latency, and personalization tradeoffs.

Questions Asked (5)

Q1

Design an ML system that predicts and ranks destination suggestions in the Uber app as a user types the first few characters into the search bar.

System DesignTechnical Trade-offsData Modeling
Author's notes

This is the whole question, basically.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying requirements and constraints (latency, scale, personalization) and framing the problem as a low-latency ranking task with candidate generation. Then walk through the ML system design end-to-end: data sources, feature engineering, model architecture, serving infrastructure, and evaluation, emphasizing trade-offs at each stage.

Pro tip: Emphasize the two-stage architecture (candidate generation + ranking) and discuss how you'd handle cold-start and real-time personalization without adding latency—this shows you understand production ML at Uber's scale.

1. Clarify requirements and scope

Ask about latency targets (e.g., <100ms), scale (QPS, number of users), personalization needs, and data availability. Define success metrics like CTR, conversion, or user satisfaction.

2. Design data pipeline and features

Identify data sources: historical trips, user profiles, real-time context (location, time), and popular destinations. Engineer features like prefix embeddings, user history, geospatial context, and popularity signals.

3. Propose model architecture

Outline a two-stage system: candidate generation (e.g., using prefix-based retrieval or ANN) followed by a ranking model (e.g., gradient-boosted trees or neural network) that scores candidates using features.

4. Address serving and latency

Explain how to serve predictions in real-time: precompute candidate sets, use caching, optimize model inference, and handle dynamic updates. Discuss fallback strategies for low-latency guarantees.

5. Plan evaluation and iteration

Define offline metrics (NDCG, recall@k) and online A/B testing. Discuss monitoring, feedback loops, and how to handle cold-start and data drift.

Key Points to Mention

  • Two-stage architecture: candidate generation + ranking to balance latency and relevance
  • Feature engineering: prefix embeddings, user history, geospatial context, and real-time signals
  • Handling cold-start for new users or destinations via popularity and context-based fallbacks
  • Latency optimization: caching, precomputation, model distillation, and efficient serving
  • Evaluation strategy: offline metrics (NDCG, recall) and online A/B testing with guardrail metrics
  • Scalability considerations: distributed training, real-time feature stores, and model updates

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

Q2

What features would you use in the ranking model, and would you go with a gradient-boosted model or a neural ranker?

Technical Trade-offsSystem Design
Author's notes

Went with gradient boosting first and justified it on latency grounds since you're re-ranking on every keystroke under 50ms.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the problem in terms of Uber's ranking context (e.g., rider-driver matching, food delivery ranking) and the trade-offs between model complexity, latency, and interpretability. Then describe a feature set covering user, driver/item, context, and interaction features, and justify your model choice by comparing gradient-boosted trees (GBDT) and neural rankers on axes like data volume, feature types, latency constraints, and ease of iteration. Conclude with a pragmatic recommendation (often a hybrid or staged approach) and mention how you'd validate the choice with offline metrics and online A/B tests.

Pro tip: Emphasize that at Uber's scale, latency and infrastructure constraints often dominate model choice—showing you'd start with a strong GBDT baseline and only move to a neural ranker when you have sufficient data and a clear need for representation learning or multi-task objectives demonstrates production maturity.

1. Clarify the ranking context and objectives

Ask or state assumptions about what is being ranked (e.g., drivers for a rider, restaurants for a user), the business objective (conversion, ETA, satisfaction), and constraints (latency, throughput). This grounds the feature and model discussion in a concrete problem.

2. Propose a comprehensive feature set

Organize features into categories: user features (history, preferences), item/driver features (ratings, acceptance rate, location), context features (time, weather, traffic), and interaction features (user-item affinity, distance, past interactions). Mention both real-time and batch-computed features.

3. Compare GBDT vs. neural ranker on key trade-offs

Discuss GBDT strengths: fast training, interpretability, strong performance on tabular data, low latency. Neural ranker strengths: representation learning for high-cardinality IDs, multi-task learning, flexibility with unstructured data. Highlight data volume, feature engineering effort, and serving infrastructure as deciding factors.

4. Recommend a pragmatic approach

Suggest starting with a GBDT baseline (e.g., XGBoost/LightGBM) for quick iteration and interpretability, then evaluate a neural ranker (e.g., two-tower or deep & cross) if data scale and latency allow. Mention hybrid approaches like using GBDT for feature transformation or ensembling.

5. Outline validation and iteration plan

Describe offline evaluation (NDCG, AUC, calibration) and online A/B testing with guardrail metrics (latency, revenue, user satisfaction). Emphasize monitoring and retraining cadence to ensure the model stays effective.

Key Points to Mention

  • Feature categories: user, item/driver, context, and interaction features, with examples relevant to Uber (e.g., ETA, driver acceptance rate, rider history).
  • GBDT advantages: strong tabular performance, fast training, interpretability, low inference latency, and ease of deployment.
  • Neural ranker advantages: learned embeddings for high-cardinality features, multi-task learning, and ability to incorporate unstructured data (e.g., text, images).
  • Trade-offs: data volume requirements, feature engineering overhead, serving infrastructure complexity, and latency constraints at Uber's scale.
  • Hybrid or staged approach: start with GBDT baseline, then explore neural rankers when justified by data and business needs.
  • Evaluation strategy: offline metrics (NDCG, AUC) and online A/B tests with guardrail metrics like latency and user satisfaction.

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

Q3

How would you handle cold-start for new users who have no trip history?

Product Sense & IdeationTechnical Trade-offs
Author's notes

Pretty short exchange.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the specific cold-start scenario (e.g., new user with no trip history) and the business objective (e.g., improving match rate or ETA prediction). Then propose a multi-pronged solution that leverages available signals (contextual, demographic, cross-domain) and fallback strategies, while discussing trade-offs and evaluation metrics.

Pro tip: Emphasize the importance of a fast-feedback loop: use online learning or bandits to quickly adapt as the user generates their first few trips, and design the system to gracefully degrade to simpler heuristics when data is sparse.

1. Clarify the problem and constraints

Ask clarifying questions to understand which model(s) are affected (e.g., ETA, matching, pricing), what data is available at signup, and what business metrics matter most.

2. Identify available signals

List potential proxies for user preference: device type, location, time of day, signup channel, and any cross-domain data (e.g., Uber Eats history).

3. Propose modeling approaches

Suggest techniques like hierarchical models, meta-learning, or feature-based cold-start (e.g., using user demographics) and fallback to global/popularity-based models.

4. Design fallback and exploration strategy

Outline how to handle uncertainty: use bandits for exploration, default to heuristics, and gradually personalize as data accumulates.

5. Define evaluation and iteration plan

Specify offline metrics (e.g., coverage, precision) and online A/B tests, and describe how to measure improvement over time as users generate history.

Key Points to Mention

  • Use of contextual features (location, time, device) as proxies for user intent
  • Cross-domain transfer learning (e.g., leveraging Uber Eats data if available)
  • Hierarchical Bayesian models or meta-learning to share statistical strength across users
  • Exploration-exploitation trade-off via multi-armed bandits for new users
  • Fallback to non-personalized models (e.g., global average ETA) to ensure baseline performance
  • Online learning to quickly adapt as the user generates trip history

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

Q4

What offline and online metrics would you use to evaluate this system, and how do they relate to each other?

Product Analytics & MetricsA/B Testing & Experimentation
Author's notes

I led with MRR and hit-at-k for offline, then CTR on suggestions and a 'characters saved' metric for online.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the system's purpose and defining success from both user and business perspectives. Then, propose a balanced set of offline and online metrics, explaining how offline metrics serve as proxies for online outcomes and how online metrics validate offline findings. Emphasize the iterative relationship between the two.

Pro tip: Highlight the importance of guardrail metrics to catch unintended consequences, and mention how offline metrics can be leading indicators for online metrics, but always validate with online experiments.

1. Clarify system goals and context

Ask clarifying questions to understand the system's purpose, target users, and business objectives. This ensures metrics align with actual goals.

2. Define offline metrics

Propose offline metrics that measure model performance (e.g., precision, recall, AUC) and are computable from historical data. Explain their role as proxies for online performance.

3. Define online metrics

Propose online metrics that reflect real-world impact, such as user engagement, conversion rates, and business KPIs. Include guardrail metrics to monitor for negative side effects.

4. Explain the relationship

Describe how offline metrics can predict online outcomes, but note that correlation must be validated through A/B testing. Discuss how online metrics provide feedback to refine offline metrics.

5. Address trade-offs and iteration

Discuss potential trade-offs between offline and online metrics (e.g., optimizing for offline accuracy might hurt user experience) and how to iterate based on both.

Key Points to Mention

  • Offline metrics: precision, recall, F1, AUC, RMSE, etc., depending on the task.
  • Online metrics: click-through rate, conversion rate, user engagement, retention, revenue.
  • Guardrail metrics: latency, error rates, fairness metrics, to ensure no harm.
  • A/B testing as the gold standard to validate offline metrics and measure causal impact.
  • Leading vs. lagging indicators: offline metrics as leading indicators for online outcomes.
  • Feedback loop: online performance informs offline model retraining and metric selection.

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

Q5

How do you think about the privacy and personalization tradeoff in a system like this?

Technical Trade-offsProduct Strategy
Author's notes

Didn't see this one coming as a serious discussion point.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing privacy and personalization as a spectrum rather than a binary tradeoff, emphasizing that the right balance depends on user trust, product context, and regulatory constraints. Then walk through a structured decision process: define the personalization goal, assess privacy risks, choose appropriate techniques (e.g., differential privacy, federated learning), and measure outcomes. Conclude with how you would operationalize this at Uber, highlighting user control and transparency.

Pro tip: Emphasize that privacy-preserving ML can often improve personalization by increasing user trust and data sharing willingness—frame it as a win-win when done right, not just a compliance cost.

1. Clarify the personalization objective and data needs

Identify what user data is truly necessary for the personalization feature and whether less sensitive or aggregated data could achieve similar results.

2. Assess privacy risks and regulatory context

Map data flows, identify sensitive attributes, and consider regulations (GDPR, CCPA) and Uber's internal privacy policies to quantify risk.

3. Select privacy-enhancing techniques

Choose appropriate methods such as differential privacy, federated learning, anonymization, or on-device processing based on the risk-benefit analysis.

4. Design for user control and transparency

Incorporate user consent, opt-outs, and clear explanations of how data is used to build trust and meet ethical standards.

5. Measure and iterate on the tradeoff

Define metrics for both personalization quality and privacy preservation, then A/B test and monitor to find the optimal balance over time.

Key Points to Mention

  • Differential privacy and its role in protecting individual data while enabling aggregate insights
  • Federated learning or on-device personalization to keep sensitive data local
  • User consent, transparency, and control mechanisms (e.g., privacy settings, explanations)
  • Regulatory compliance (GDPR, CCPA) and Uber's specific privacy commitments
  • Business impact: how privacy-preserving approaches can increase trust and long-term engagement
  • Metrics for evaluating the tradeoff, such as personalization lift vs. privacy budget (epsilon)

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