Track 3: Public Safety

Inspiration

In emergencies, official alerts can be confusing, delayed, or hard to understand for the general public. We wanted to create a platform where communities receive clear, fast, and reliable safety information, keeping everyone safer and calmer during critical situations. Combining official feeds, community reports, and GPT-4’s natural language capabilities, we aimed to bridge the gap between complex alerts and public understanding.

What it does

Neighborhood Safety Alert Network collects emergency alerts from official sources (police, weather, earthquake feeds) and community submissions. It uses GPT-4 to instantly simplify complex alerts into easy-to-understand updates and allows users to ask follow-up questions for clarification. Users subscribe to alerts by their area and receive notifications via a real-time web app, ensuring they stay informed, safe, and connected(future).

How we built it

Frontend: React + TypeScript built with Vite for fast performance. Styling: TailwindCSS and custom reusable UI components. Backend (planned): Flask for data aggregation and GPT integration. Alert Integration(planned): Simulated real-time alerts from NOAA, USGS, and sample police reports. Hosting: Deployed on Netlify for fast, reliable access. AI Processing: GPT-4 to summarize, rephrase, and answer follow-up questions about alerts.

Challenges we ran into

Designing a clear yet flexible UI for different emergencies (crime, fire, weather, etc.). Ensuring GPT responses remain factual and cautious, especially when details are still emerging.

Accomplishments that we're proud of

Built a real-time alert network prototype in under 24 hours. Successfully integrated GPT-4 for both summarization and contextual Q&A. Designed a clean, scalable architecture that can grow into a full production system. Created an app that can truly improve public safety and community trust.

What we learned

How to design with emergency communication principles (clarity, speed, responsibility). Best practices for integrating AI models into real-time systems. Balancing technical features with user experience, especially for high-stress scenarios.

What's next for Neighborhood Safety Alert Network

Full backend integration to automate real-time ingestion of official and community alerts. Push notifications via SMS/email using AWS SNS for critical incidents. Bilingual alert support using GPT-4 or AWS Translate. Advanced trust scores for community-submitted alerts using AI moderation. Scalable rollout to pilot neighborhoods and city partnerships.

Built With

Share this project:

Updates