Privalino protects kids from cyber dangers. Using deep learning methods, Privalinos AI is able to detect these dangers early on and alarm the children, their parents or our human moderators.
We treat the problem as an NLP challenge: Our recurrent neural networks read chat logs from our messaging app and decide whether a danger is present. If a certain threshold is passed, we ring the alarm. Human moderators can review and annotate cases in an internal web dashboard while parents can give feedback