ZZomato·BehavioralML-2System Design

Design a Dish Recommender System for Zomato (Case Study)

viaGlassdoor

Case Study: How would you design a recommender system for Zomato to recommend dishes to a user? Approach: Clarify objective (engagement vs order conversion), then discuss candidate generation (collaborative filtering on order history, content-based signals from dish/cuisine attributes, popularity/trending signals) and ranking (a model blending relevance, freshness, price sensitivity, restaurant availability). Cover data pipeline, feature freshness, and how to A/B test recommendation quality against a baseline (e.g. popularity-based).

Add a follow-up question they asked
Handling the Cold-Start Problem
Choosing an Evaluation Metric for the Recommender
asked …
LeaderboardSalary
Language
Account