YOUR BRIEF
What to work through
Work through a recommendation system as a spoken system design interview. Start by clarifying the scope, sketch a working design and explain the trade-offs behind your choices.
- Show each profile rows of recommended titles on the home page
- Update the picks as the viewer watches
- Record viewing signals (plays, hovers) to learn from
- Show something sensible to a new viewer with no history
BEFORE YOU START
Constraints to discuss
- Meta-scale traffic: millions of requests a second
- Candidates narrowed in stages, from over 10,000 down to about 50
- 5% to 10% of slots kept for exploration
- The home page still loads when a dependency fails
Need a starting point?
Watch for these common gaps as you prepare:
- Running the heavy model over the whole catalogue on each request
- No fallback, so a ranking outage leaves an empty home page
- Features computed one way for training and another for serving
- Treating every signal as equally reliable
- Not saying which team owns which part when several teams share the pipeline
Why this question is in the collection
These sources connect the problem to company interview reports or preparation guides. They do not guarantee what a company will ask you.
- Netflix: Aced’s Netflix guide reports the prompt of a dynamic recommendation system for the Netflix home page, with an interviewer probing hover as a signal.
- Meta: Aced’s Meta guide lists a product recommendation system, personalized news ranking and Instagram’s Explore page among its ML design questions.