Online bullying detector
AI that spots bullying before it hurts
01
Problem
EVERY THIRD CHILD IS A VICTIM OF A DIGITAL VIOLENCE.
EVERY THIRD CHILD WHO IS A VICTIM NEVER TELLS ANYONE, NOT EVEN THEIR PARENTS.
Children are increasingly exposed to online threats and Parents often lack time & visibility into what happens. The consequences are devastating:
- emotional distress
- mental health struggles
- FATAL
02
Solution
CyberParent is a mobile app installed on a child’s phone. It uses AI to scan children's messages for signs of bullying in real-time and alerts parents, enabling them early intervention before any damage is done.
03
Unfair advantage
Our AI system combines parents’ reports, cyberbullying patterns among children in the same environment, and each child’s individual profile. This way, our AI is personalized for every child.
04
Everything runs locally
We created a light AI model and everything runs locally on the child's phone — data is not leaving the phone, no servers. We don't store any sensitive data.
05
We are app-agnostic
Our solution doesn't depend on social media and we can easily add new chat apps into our app portfolio.
06
Traction
Our app is working in:
- 1 public school in Mostar, Bosnia (200 parents)
- 1 private school in Mostar, Bosnia (60 parents)
- 3 LOI with private schools in Belgrade, Serbia (in progress)
- 1 LOI from child organisation in London (in progress)
07
How CyberParent helps parents:
Timely warnings
Our app sends an immediate warning, allowing parents to react in a timely manner and protect their child.
Saves time and reduces conflict
Parents don't have to manually review all messages, the app automatically notifies them of anything risky. This saves them valuable time and avoids unpleasant arguments with their children.
Safety and privacy for children
The app only sends suspicious messages to the parent, not everything.
08
Team
Help us build safer digital environment for families and be part of unique incremental-learning AI project.
Let's explore potential collaboration.
Unicorn in training 🦄
09