Indiana Public Safety Assistant

A conversational assistant for reporting emergencies, accessing safety info, and connecting with Indiana’s public resources.

Comment

🛠️ Project Story: Indiana Public Safety Assistant

🔍 About the Project

The Indiana Public Safety Assistant was born from a desire to leverage AI in a meaningful, community-focused way. Inspired by the growing need for accessible safety information, especially in underserved or rural areas of Indiana, this chatbot was created to act as a real-time, responsive public safety companion.

The project aims to bridge the gap between citizens and the safety resources they need — combining modern tools like Groq’s fast AI inference and DuckDuckGo's privacy-focused search with a clean and approachable UI powered by Streamlit.

💡 Inspiration

Emergencies can happen at any time, and not everyone knows who to call, where to go, or what steps to take. During recent natural events in the Midwest (like tornado warnings and flood alerts), it became clear how valuable a simple, centralized tool could be — especially one that speaks in plain language, reacts quickly, and offers credible, localized information.

🏗️ How It Was Built

  • Frontend/UI: Built using Streamlit for a lightweight, interactive interface.
  • LLM Integration: Powered by Groq API, offering fast and context-aware responses.
  • Web Search: Enhanced with DuckDuckGo search to surface real-time public safety updates.
  • Data Sources:
    • JSON files for local emergency resources and hazard reports
    • Real-time weather alerts via APIs (optional/future enhancement)
  • Hosting: Designed to be deployed on Streamlit Cloud with secure secrets management.

🚀 Key Features

  • Hazard and emergency reporting
  • Public safety tips and guidelines
  • Instant access to emergency contact numbers
  • Weather alerts and resource links
  • DuckDuckGo-enhanced smart search
  • Customizable and open for community contributions

🤯 Challenges Faced

  • Latency: Ensuring real-time responsiveness while integrating multiple APIs.
  • Context Handling: Managing user prompts so the assistant can stay relevant across multi-turn conversations.
  • Security: Safely handling API keys and user-generated content without exposing sensitive data.
  • Local Data: Collecting and formatting community-specific emergency resources for Indiana.

📚 What I Learned

  • How to deploy and scale AI-powered apps with Streamlit Cloud
  • Best practices for working with LLMs and external APIs
  • The value of clear UX design in emergency tools
  • Importance of modular code structure and secrets management in production-ready applications

🌱 Future Plans

  • Add voice input for hands-free emergency assistance
  • Integrate official weather and emergency alert APIs (like FEMA or NOAA)
  • Support for multiple languages
  • Broaden resource data for other U.S. states

🙌 Final Thoughts

This project reinforced the power of combining simple tools with smart AI to solve real problems. The Indiana Public Safety Assistant is just the beginning — and it’s open to contributions, feedback, and expansion to help make safety accessible for everyone.

Built With

Share this project:

Updates