The problem Sentinel : An AI that watches your back solves

It is a big challenge for the online moderator to go through all the comments and flag (or delete) certain offensive and unsavoury comments aimed at individuals or against the video. We provide a robust AI algorithm to automatically identify such potential comments in real time. The project is implemented in both Web and Android Platform.

Challenges we ran into

Modelling AI algorithms is by far the most toughest challenge we came across. We modelled our own Deep Learning model right from scratch to classify comments with good and bad sentiments. We are able to achieve an accuracy of around 85% which is quite around the state of the art results so far.