Today, when we deal with pressing crises as a global society like COVID-19 right now, we use the Internet as a necessary form of communication, and often our only one. However, the most vulnerable members of our society struggle with reliable Internet access. 27% of Americans over 65 years old don’t use the Internet at all, but this is the group most susceptible to COVID-19. We need a bridge between innovative breakthroughs in communication technology and legacy systems of the past that allows vulnerable demographics to stay as up-to-date on developing crises as possible.
What it does
COVcall seeks to bring the instant-update experience of social media platforms like Twitter to elderly citizens in a digestible, familiar format: a classic phone call. By referencing timelines of select verified individuals like State Governors, COVcall aggregates information relevant to each state’s individuals and automates an outgoing call that reads out-loud necessary information. This ensures that COVcall not only updates senior citizens about national changes, but also provides new information in regional contexts.
How we built it
COVcall uses the Twilio API to set up automated voice calls. People can call our number at +14073293385 to sign up. On the backend, COVcall uses a Flask server to manage these inbound calls. When they sign up, the Flask server adds their phone number and state to our MongoDB database. Next, COVcall uses BeautifulSoup, a Python web-scraping library, to scrape state governors' twitter feeds. When it scrapes a relevant tweet, it uses Twilio to call everyone from that governor’s state and reads the tweet aloud over the phone.
Challenges we ran into
Time crunch was the biggest issue we had. We had so many cool ideas about refining the basic process of analyzing text and automating calls (see below!) that we’re sure to keep working past this hackathon.
What we learned
We learned a lot of new things: implementing a SQLite database in Python practicing the fundamentals of web scraping and parsing text using Figma as a UI/UX tool gaining experience with Twilio’s API for the first time
What's next for COVcall
Implementing an NLP algorithm to aggregate and distinguish relevant tweets from irrelevant ones Give users multiple options for how/when they can receive an update Improve the database query times by implementing a cache and other optimization techniques, including exploring database management altogether Integrating with local governments, retirement centers, and community organizations to reach the citizens