Case Study: NSFW achieves faster, lower-cost content moderation with Nanonets NSFW API

NSFW - Customer Case Study

NSFW, a social website handling more than 50 million image uploads per month, needed an efficient way to moderate objectionable content at massive scale. Manual review was too slow, expensive, inconsistent, and risky for human moderators, so the company turned to Nanonets’ NSFW content moderation model.

Nanonets built a custom image classification model for NSFW and SFW using 20,000 training images per class and deployed it on the customer’s infrastructure via Docker. The solution increased review speed from 800 images per human per hour to 20,000 images per hour and reduced the cost of reviewing images from $38 to $1, while enabling automated, customer-specific moderation.