System designHardsd-g340

Subject RecommendationLevel Senior–Staff~45 minCommon in ML systems interviewsIndustries Technology

Question

Design the cold-start strategy for a recommendation system on two fronts at once: brand-new *users* (no history) and brand-new *items* (no engagement data, e.g. a just-published article or a just-listed product). The collaborative-filtering / engagement-trained ranker is useless for both. How do you give new users decent recommendations from minute one, how do you give new items a fair chance to be discovered without flooding everyone with untested content, and how do you stop the rich-get-richer feedback loop where popular items always win?

What a strong answer looks like

Clarify scale and constraints first. Propose a clean component breakdown, then go deep on the hard parts — data model, bottlenecks, consistency, failure modes — and name the trade-offs you are making.

Learn the concepts

Narrate your design

Loading whiteboard…

Run or narrate your approach, then ask the coach.