KinderGuardians

Logo
Report
Child_Home
Parent_Login
Create_Account

Inspiration

We were deeply moved by a case involving an 8-year-old girl who tragically became the victim of a horrific sexual assault. It broke our heart to realize that while we have many safeguarding tools for women, there are limited protection devices for toddlers who cannot defend themselves or signal for help. So KinderGuardians seeks to serve the most vulnerable—toddlers—by recognizing red-flag language.

What it does

Child safety remains a critical concern, especially for toddlers who are often unable to articulate or defend themselves against potential abuse. Our project addresses this issue by using technology to identify and alert caregivers about a potential sexual harrasment, using machine learning and natural language processing, KinderGuardians can detect phrases that are often used by sexual predators to manipulate children.

How we built it

Our methodology includes fine-tuning an LLM using the synthetic dataset. We integrated this trained model into a web application featuring an intuitive interface. The web app enables voice recordings that are processed by the model to classify statements as red or green flags. Additionally, a parent-only login feature allows secure access and control.

Accomplishments that we're proud of

We bein able to curate relavant data and being able to train the model. We understand that there is quite some work that we were unable to finish yet but this accomplishment in one day is something that we are proud about. we hope this projects takes life and safeguards as many children as poswsible.

What's next for KinderGuardians

KinderGuardians is designed as a web application to classify phrases as “red flag” (potentially harmful) or “green flag” (harmless). Future expansions may include hardware components to enable IoT integrations, such as wearable devices that passively monitor the child’s environment. For the hackathon, our focus is on developing a software prototype that alerts parents or guardians to potential risks.

Conclusion

KinderGuardians is a significant step towards providing proactive protection for children. Early intervention by parents, enabled by this tool, can safeguard kids from potential abuse, reduce the likelihood of major assaults, and help prevent long-term trauma. While the prototype is currently limited to software, it demonstrates the potential for using AI to identify harmful language patterns and alert caregivers in real-time. With further development, KinderGuardians could evolve into a robust, IoT-enabled tool for comprehensive child protection.