We wanted to address the lack of infrastructure caused by natural disasters in developing countries, such as the recent hurricanes and tsunamis. After being ravaged by these natural disasters, many of these areas do not have widespread or reliable access to the internet, and many citizens only own flip phones that can call and text. In order to help distribute necessary emergency supplies and need in an efficient manner, we decided to build Beaver Helpline.
What it does
Beaver Helpline provides an automated phone number that records calls giving locations, names, and needs of individuals using Google Voice, and converts it through a pipeline into text primarily through Rev.ai. The transcript is processed using NLP techniques in Google Cloud to extract the relevant information - locations, names, and needs. These locations are then converted into coordinates and are plotted on Google maps with the relevant information for efficient distribution and logistical planning. We have also implemented a Slack notification system using IBM AppConnect and storage for all of the resulting .mp3s in an easily accessible Dropbox folder.
What we learned
We've learned a lot about integrating different APIs and trying to form a cohesive project that uses many services.
Call (781) 613-3345 to try it out. Repository is available at https://bitbucket.org/dbli2000/hackmit2018/src/master/