We have all felt the subtle panic when Google tells us "you are coughing because you have cancer". In attempt to eliminate that, we built a diagnosing software that is easily accessible, through Kik, and uses real medical records and ML to give a better diagnosis than a simple Google search.
What it does
Receives text input of symptoms, processes it with NLP, and feeds it through an ML algorithm to output possible diagnosis and nearby medical locations.
How I built it
Python kik Firebase/Pyrebase Google-place Google-maps Fuzzyset Flask Apimedic Challenges I ran into Apimedic API was hard to implement and we had to deal with a call time limit Attempted to implement custom keyboard on Kik but had little time to finish Accomplishments that I'm proud of No team member knew all of the APIs we implemented but we were very efficient in on-site learning. Made a functional app in less than 12 hours
What I learned
How to create a Flask Kik app How to implement NLP with Fuzzyset How to get diagnosis from symptoms using apimedic
What's next for BPatient
Give users more freedom in entering relevant data Receive health data from devices such as AppleWatch and Fitbit Onsite automatic report to medical professionals if the condition is considered severe