Design a Recommendation & Feed Ranking System

Advanced · 28 min read · High-Level Design, ML & AI Systems

Candidate generation, two-tower retrieval, multi-task ranking, re-ranking, training data and cold start.

This is an Advanced lesson, part of Premium. What it covers:

• 1. Clarify the problem
• 2. Frame it as an ML problem
• 3. Metrics
• 4. Capacity estimation
• 5. High-level design
• 6. Deep dives
• 7. Failure modes
• 8. Trade-offs summary
• 9. What interviewers look for
• Interview questions
• Practice
• Further reading

Previous: The ML System Design Interview · Next: Design Search Ranking