After Hurricane Maria hit Puerto Rico on October of 2017 the cell towers went down and there was no internet. The only mode of communication was by SMS.

Information is critical to making decisions. In developing countries or emergency situations, connecting with healthcare services, different local and regional markets, or community bulletins can be challenging. Although smartphone coverage remains low, approximately 70% of people in developing countries use cell phones for SMS. Our goal is to provide the essential information which can make people's lives easier.

What it does

WNet uses SMS to connect users with their communities, markets, governments, and international organizations. It provides user generated bulletins, market listings, and critical news information.

How we built it

The chatbot portion was built using DialogFlow and Twilio’s Programmable SMS API. It relies on a MySQL database hosted in AWS as well as a Python Flask backend.

Challenges we ran into

We had no previous knowledge of MySQL, AWS, Chatbots (Twilio), and DialogFlow. As such, we had difficulty first getting started with the project, but as time went on, the project began to come together.

Accomplishments that we're proud of

Our team had never built a program with a backend before, so we took this opportunity to learn about hosting, databases, chatbots, artificial intelligence, webapps, and how they all fit together. We are also proud of finding time to sleep during the Hackathon.

What we learned

We learned how to create SMS Chatbots, how to use DialogFlow, how to host databases on AWS, and how to organize a MySQL database.

What's next for WNET

We would like to create more channels for local authorities. Moreover, WNet would evolve to provide the ability to better connect users with international organizations for help and emergency notices. If we gained enough resources and support, we would also like to add many more languages for support in the SMS chatbot so that it would become more accessible to everyone.

Share this project: