Learning objectives
- • Design candidate generation, ranking, reranking, and exploration layers
- • Handle cold start, position bias, diversity, and delayed labels
- • Connect offline ranking metrics to online product outcomes
System Design
Retrieval, Ranking, and Recommendation Systems
Prepare multi-stage recommender and search designs with retrieval, ranking, reranking, diversity, freshness, and feedback loops.
Recommended on day 63120 minutesAdvanced
Interview prompts
- • How do you design a short-video recommender end to end?
- • How do you handle position bias in click logs?
Prerequisites