← Grab Interview Insights

Grab·Machine Learning Engineer·Technical Phone Screen·Intermediate

Intermediate
Jun 2026

Summary

Grab ML engineer round that's basically a live Kaggle competition. You get a link, a dataset, and the task is to predict order fulfillment time. An AI agent can write the code but you're expected to drive all the decisions around features, model choice, and evaluation.

Questions Asked (2)

Q1

Given a dataset of food delivery orders, how would you formulate predicting order fulfillment time as an ML problem, and what features would you engineer from the available signals?

Data ModelingTechnical Trade-offs
Author's notes

This is where most of the interview actually lives.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the problem as a supervised regression task predicting order fulfillment time, then discuss feature engineering from multiple data sources. Emphasize the importance of defining the target variable precisely and considering real-world constraints like data leakage and temporal aspects.

Pro tip: Mention that you would use a quantile loss or predict a distribution rather than a point estimate, because delivery time estimates are often communicated as ranges to customers, and this also helps with uncertainty quantification.

1. Define the ML Problem

Clarify the prediction target: order fulfillment time from order placement to delivery. Decide between regression (predicting exact time) or quantile regression (predicting percentiles) based on business needs.

2. Identify Data Sources and Signals

List available data: order details, restaurant prep times, driver location and availability, traffic conditions, weather, historical delivery times, and customer location.

3. Engineer Features

Create features such as distance between restaurant and customer, estimated prep time, driver-to-restaurant distance, time of day, day of week, weather conditions, and historical average delivery time for similar orders.

4. Handle Temporal and Leakage Issues

Ensure features are available at prediction time (e.g., no future information). Use time-based splits for validation to mimic real-world deployment.

5. Evaluate and Iterate

Choose appropriate metrics (MAE, RMSE, quantile loss) and validate with business metrics. Consider model interpretability and potential biases.

Key Points to Mention

  • Target variable definition: order fulfillment time from placement to delivery, possibly excluding cancellations.
  • Feature engineering from multiple sources: order, restaurant, driver, traffic, weather, and historical data.
  • Temporal validation to prevent data leakage and ensure model generalizes to future orders.
  • Use of quantile regression or distributional predictions to provide uncertainty estimates.
  • Handling missing data and real-time constraints (e.g., driver location updates).
  • Business impact: improving ETA accuracy to enhance customer experience and operational efficiency.

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

Q2

Which regression model would you choose for this problem and why, and how would you evaluate its performance?

Technical Trade-offsProduct Analytics & Metrics
Author's notes

Went with gradient-boosted trees, which felt safe.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the problem context—data size, feature types, and business objective—then propose a baseline model (e.g., linear regression) and justify moving to a more complex model if needed. Explain your choice based on trade-offs between interpretability, performance, and scalability, and outline a robust evaluation strategy using appropriate metrics and validation techniques.

Pro tip: Always tie your model choice back to business impact—at Grab, this means considering operational constraints like latency, explainability for stakeholders, and how the model will be deployed. Mention that you'd start simple and iterate, showing you value pragmatism over complexity.

1. Clarify the Problem and Data

Ask about the dataset size, feature types (categorical, numerical, text), target variable distribution, and business goal (e.g., predicting delivery time, customer lifetime value). This ensures your model choice aligns with the problem.

2. Propose a Baseline and Justify Model Selection

Start with a simple, interpretable model like linear regression (or regularized variants) as a baseline. Then, if needed, suggest more complex models (e.g., gradient boosting, random forest) and explain the trade-offs in terms of accuracy, interpretability, training time, and inference latency.

3. Outline Evaluation Metrics

Choose metrics based on the problem: RMSE/MAE for regression, but also consider business-specific metrics (e.g., percentage of predictions within a tolerable error). Mention the importance of using a hold-out test set and cross-validation.

4. Discuss Validation Strategy and Potential Pitfalls

Explain how you would split the data (time-based split if temporal), handle overfitting, and check for data leakage. Mention techniques like k-fold cross-validation and monitoring for concept drift.

5. Connect to Deployment and Monitoring

Briefly discuss how the model will be deployed (batch vs. real-time) and monitored post-deployment, including retraining frequency and performance degradation checks.

Key Points to Mention

  • Trade-offs between interpretability (linear models) and predictive power (ensemble methods)
  • Regularization techniques (Lasso, Ridge) to handle overfitting and feature selection
  • Evaluation metrics: RMSE, MAE, R-squared, and business-specific metrics
  • Cross-validation strategies, especially for time-series data (e.g., time-series split)
  • Handling of categorical variables (one-hot encoding, target encoding) and missing data
  • Scalability and inference latency considerations for production

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