Toxic speech and hatred are tarnishing the value of online community spaces. Studies have shown that prolonged exposure to hateful content and messages affects your mental state.
We trained a multilingual toxicity classification model to filter out toxic messages or conversations.
We built our product to scale so we trained and deployed the model in AWS with a load balancer to manage a high volume of traffic.
Tackle cyber-bullying in online forums and channels
Help moderators to maintain a safe community space to have conversations