Design a Food Delivery Recommender
Company: Uber
Role: Machine Learning Engineer
Category: ML System Design
Difficulty: medium
Interview Round: Onsite
Design a recommendation system for a food delivery app similar to Uber Eats.
When a user opens the home page, the system should recommend restaurants or dishes that the user is most likely to order. The design should personalize results using factors such as past orders, cuisine preferences, time of day, delivery address, restaurant availability, estimated delivery time, and marketplace conditions.
Discuss:
- Product goals and success metrics
- Candidate generation and ranking
- Features and training data
- Cold-start handling for new users and new restaurants
- Online serving architecture and latency constraints
- Exploration, diversity, and business rules
- Offline and online evaluation
- Common failure modes and how to mitigate them
Quick Answer: This question evaluates competency in designing personalized, scalable recommendation systems, testing knowledge of candidate generation and ranking, feature engineering and training data, cold-start strategies, online serving and latency constraints, exploration/diversity trade-offs, business rules, and offline/online evaluation.