Inspiration

Over 60 million children are active internet users, which presents a unique challenge as harmful content continues to spread online. With over 3 million instances of such content, much of it remains easily accessible and undetected, making it a growing concern for parents. Research showed a clear need for a solution to help protect children from harmful content online. This insight led us to create Snailly, an extension that helps protect children while giving parents control over their kids' online experiences.

What it does

Snailly is a browser extension that blocks harmful website content using Gemini Nano, ensuring a safer browsing environment for children. The extension also includes a summary feature that allows for quick understanding of a webpage's content, so parents and children can easily understand the content on the page. The extension specifically targets harmful content, classifying it as “Positive” or “Negative” based on content analysis. When negative content is detected, Snailly prevents access by blocking it, promoting a safer and more educational online experience.

How we built it

We built Snailly as a browser extension using React and Vite for efficient performance, styled with TailwindCSS and Shadcn for a modern, responsive interface. The extension utilizes the Gemini Nano Model (Prompt API) for content classification, analyzing websites in real-time to categorize them as "Positive" (safe) or "Negative" (harmful), with unsafe content automatically blocked. Additionally, we integrated the Summarization API to generate concise summaries of webpage content, helping parents and children quickly understand a page content.

Challenges we ran into

One of the main challenges was ensuring that the Gemini Nano model could accurately and consistently classify content from multiple sources. Sometimes the results of the Gemini Nano model did not match the predictions, although it often showed the right response. Balancing a user-friendly interface with robust functionality was another challenge, as we wanted to ensure Snailly was both intuitive for parents and effective in blocking harmful content.

Accomplishments that we're proud of

A proud achievement, we are proud to have successfully integrated the Gemini Nano model to classify web content and block malicious sites in real-time that can run in multiple languages. The summary feature is also a unique addition, allowing parents and children to quickly understand the content of a page without having to browse it in its entirety. Overall, this extension provides a practical solution to the pressing issue of online child safety.

What we learned

Through this project, we learned how to apply AI to real-world challenges, especially when dealing with potentially harmful content online. We gained hands-on experience in implementing content classification models, integrating Gemini Nano such us PromptAPI and SummarizationAPI, and optimizing browser extensions for seamless user experiences.

What's next for Snailly - Safe Browsing For The Children

In the future, we plan to expand Snailly's functionality by adding more customizable controls for parents, such as blocking harmful images and videos containing inappropriate audio with harsh language unsuitable for children. Additionally, we aim to integrate with a server-side API to enable real-time updates and centralized data management, enhancing overall performance and scalability. To further engage users, we plan to introduce gamification features, such as reward systems for safe browsing habits, encouraging children to learn and explore responsibly.

Built With

Share this project:

Updates