Note: This project won AngelHack 2019 Silicon Valley's IBM Call for Code Global Challenge prize


We knew that when disaster strikes, people have to move—fast. Search-and-rescue teams, first responders, and humanitarian organizations need information in real-time about where they can and can not go. With Alleviair, that information can be provided by anyone and everyone, seamlessly shared across platforms.

What it does

Individuals who discover unusual traffic situations or emergencies can submit their concerns to our chatbot. Our machine-learning algorithm then analyzes users’ submissions to pair descriptions with easy-to-understand visualizations, accessible via Aid organizations can then draw on that shared knowledge to make better decisions about how to get where they’re going, saving time and lives.

How we built it

We used Botkit to make a Facebook Messenger chatbot hosted on Heroku that can interact with users. To analyze user messages, we used the IBM Watson Natural Language Understanding Library to extract essential information on traffic situations. After analyzation, message data is stored in a PostgreSQL database on Amazon RDS. Our web app then makes requests to the backend to populate its Google Map with disaster data (message) locations.

Challenges we ran into

On the backend side, we found the most difficulties in linking up the RDS database to our chatbot handler and reliably handling SQL queries; on the frontend side, we found it difficult having our Google Map reload in live time when new locations are identified.

Accomplishments that we're proud of

We were able to use new tools like AWS RDS and IBM-Watson's libraries to create our project.

Built With

Share this project: