Your question is Build a Recommendation System. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How would you approach building a recommendation system?
Discuss how you would define the recommendation objective, represent users and items, select between collaborative filtering and content-based methods, handle cold-start users and items, and evaluate recommendations offline. Include a practical modeling approach, data split strategy, ranking metrics, and production considerations such as latency, freshness, and feedback loops.
