Inspiration

Inspired by our personal experiences, and an effort to mitigate confusion and decentralization of resources, untimeliness for getting relevant personalized information for the greater Corpus Christi area. Other inspirations include news aggregate apps, and Live universal awareness maps.

What it does

Our application provides locally focused, disaster preparation and guidance via our integrated chatbot. Ask it about evacuation routes, flood zone information, or details about your next steps to prepare or respond. Use our integrated flood zone map with open source reporting, updating traffic camera images, and notifications to stay aware of construction, traffic, fires, or other hazards. Administrators can edit and filter the map for veracity in reporting.

How we built it

In the context of Rapid Prototyping SDLC, we used Python/Flask, and LLM generated code with web-scraping libraries and open source tools to aggregate and integrate all features as well as hosting it live via nGrok to provide a free open source web-application.

Challenges we ran into

Sourcing information and appropriately placing flood zone layers into the app. Generating appropriately readable outputs from the LLM to pdf format. Handling API request responses from multiple sources in a usable manner to place on the map and into LLM "Disaster Specialist" generated content.

Accomplishments that we're proud of

The map icons and warning nodes that function live! The way we could integrate the user provided information into the Disaster Specialist to provide personalized plans, steps, and guides for the user in a responsive chat format.

What we learned

API integration! Tailoring resources to our application when API is not viable. How to integrate and prompt LLMs in code. How to better utilized generated code for rapid and effective development.

What's next for Ready - Corpus Christi

User accounts with email or SMS notification to provide specialized alerts. LLM integrated tool calls such as increased web scraping, or geographical services to increase safety and personalization to our users. Live feeds of traffic instead of timed snapshots.

Built With

Share this project:

Updates