A content moderation model flags policy-violating posts for human review. It runs at 91% precision and 34% recall. Two thirds of violating content is getting through, and the trust-and-safety team is escalating.
The dataset is heavily imbalanced: about 0.4% of posts violate policy. The model is a fine-tuned text classifier trained on 180,000 labelled examples, of which 4,000 are positive. The team's last three attempts to improve it: a larger model, more training epochs, and class weighting: all moved recall by less than 2 points.
Human review capacity is 25,000 posts per day out of 6 million posted. The team wants to know what to actually do.
Diagnose why recall is stuck and lay out what you would try, in priority order.
Build an interview-ready technical plan: frame the problem, explain the ordered approach, name the trade-offs, and show how you would validate the outcome. The AI reviewer grades that reasoning against this problem’s rubric.
Four guided sections · works on any screen