I have been on campus for a little under a week, and already we are seeing the positive Covid-19 cases skyrocket. One of the issues that we are seeing is that the university is struggling to track down the people who have been exposed to a positive individual. In the world of big data, I felt there had to be a solution to bypass the monotony.
Additionally, I have seen COVID risk calculators put out by a number of medical groups. Specifically, this one from the Cleveland Clinic uses a number of data points to predict the risk of the likelihood of testing positive.
What it does
Using location data collected by Kochava Collective, the Automated Contact Tracer can view the trails of individuals as they go about their lives. When combined with a daily health check built into the web app, this data can be used to judge the potential exposure to coronavirus an individual has throughout their day. Scored by a mixture of visiting contaminated spaces, interacting with infected individuals, and the results of the daily health/symptom survey, the user is provided a rating of how weary they should be regarding COVID. If they have a score high enough, they are recommended to visit their nearest testing center.
How I built it
This was admittedly my first time working with anything in AWS aside from S3 and Route 53. So, to start, I had to watch a lot of videos and work through a lot of tutorials. I tried to keep as much of the project within AWS as possible to continue building skills and working on technology that was new to me.
I have a ton of detail relating to how I built this on my github wiki. Please go check it out!
Challenges I ran into
While there are a ton of resources out there, I was not always able to find answers to my questions. One of the big issues I had was in figuring out how to host the python script on AWS to be run on a schedule. Ideally, I would have loved to use Lambda functions and imbed all of the code there, but I couldn't due to the lack of some packages in Lambda. I spent a ton of time trying to figure out a workaround, but none of the 'miracle cures' perscribed to me via youtube ever did the trick...
Accomplishments that I'm proud of
I am quite proud of the fact that I started with very little experience with the AWS suite and was able to self-teach my way to proficiency in a pretty good list of tools. It all fits together quite nicely, which is making me tempted to look into studying up for an AWS certification. Would definitely be a challenge, but I am up for it!
Additionally, I am proud of the fact that I was able to build this entire thing by myself. I have never done a hackathon alone, and as such, I have always been in the backend of the team, deep in the data. This time, I had to be versatile- working from python to css... No place to hide when you are working alone!
What I learned
As I have said above, I have learned quite a bit regarding new technology, but I have also learned quite a lot about individual project management. I did not hear about this challenge until mid-July, but once I did, I began to think and plan. Over my internships, I have been learning the value of planning before coding. Too many times, I have been 5 hours into code and then find out I had missed something in my thinking. For this project, I worked to think first, then act.
What's next for Automated Contact Tracer
I truly believe this is an idea that could help institutions like my university recover quicker from the pandemic. If I am to continue with the project, I may need to draft some of my 'web-developer' friends to join me. I am a Data Scientist. I have my weaknesses. Web development is definitely (for now) one of them!!
I also need to continue thinking through the privacy and legal implications of an app like this. I know the university has a lot of power, especially when facing a pandemic, so I think they may be able to work with me on something like this. I have a contact in my schools' legal office, so I might need to reach out and have a chat.