SafetyBot CivicTech Initiative

Empowering Indianapolis with Smart, Accessible Public Safety.

Comment

Inspiration

Indianapolis faces ongoing public safety challenges, and I wanted to leverage the power of AI to make safety information and resources more accessible for everyone. Inspired by civic tech initiatives and the potential of conversational AI, I set out to create a chatbot that could help residents quickly get answers to safety questions, report hazards, and find emergency resources, all in one place.

What it does

SafetyBot is an AI-powered chatbot designed to provide Indianapolis residents with real-time public safety information, emergency guidance, and a simple way to report concerns. It answers common safety questions, connects users to local resources, and offers instant support during emergencies, all through a friendly chat interface.

How we built it

I used the Google Gemini API as the core conversational engine, integrating it into a web-based chat interface. The bot’s responses are tailored using carefully crafted prompts and example interactions, ensuring it stays focused on Indianapolis public safety topics. The project was built and tested locally, with all necessary dependencies and API integrations set up in a Linux development environment. The codebase is modular, making it easy to extend with new features or connect to additional data sources in the future.

Challenges we ran into

  • Crafting prompts that keep the bot focused on accurate, relevant public safety information.
  • Ensuring the bot’s responses are based on trustworthy, up-to-date information.
  • Learning the Gemini API, managing API keys securely, and handling error responses.
  • Since the project isn’t currently deployed, I had to simulate user interactions locally.

Accomplishments that we're proud of

  • Successfully integrated the Gemini API to power a focused, helpful public safety chatbot for Indianapolis.
  • Developed a user-friendly interface and prompt system that keeps conversations on-topic and useful.
  • Built a strong foundation for future expansion, including potential deployment or integration with city data feeds.

What we learned

  • How to use the Gemini API for building conversational AI, including best practices for prompt design and error handling.
  • The importance of clear, concise system instructions to guide AI behavior.
  • Strategies for managing API keys and handling sensitive data securely.
  • That even without deployment, a well-documented, locally tested project can be compelling for hackathons.

What's next for SafetyBot CivicTech Initiative

  • Consider deploying the chatbot to a cloud service to make it accessible to real users and gather feedback.
  • Involve community members or stakeholders to test the bot and suggest improvements.
  • Add support for more safety topics, integrate live alerts, or connect with local government data feeds.
  • Explore multi-language support and mobile-friendly interfaces to reach more residents.
Share this project:

Updates