Meta
S17
HardPremiumDesign a Trust & Safety / Content Moderation Agent at Scale
Design a trust and safety system that moderates user-generated content for a platform with millions of posts a day, using models only where judgment is needed.
ClassificationSafetyScaleCost Optimization
Key Requirements
- Catch harmful content (hate, spam, CSAM) with recall set per severity
- Keep false positives low so legitimate users are not punished
- Keep publishing fast at millions of posts a day
- Appeals with a statement of reasons for every action
- Robust against users trying to evade detection
Interviewer Follow-ups
- Q1How do you handle content that is borderline or context-dependent?
- Q2How do you keep up with new types of abuse?
- Q3How do you handle appeals efficiently at scale?
Loading...