The problem HealthCarrd solves

It is often noticed that patients find hard to keep track of their health data over time. The mounting bills and reciepts, diagnosis certificates all end up confusing the patient. This ultimately causes him to miss out on crucial health information regarding himself which could be life changing. Thus the doctor is also not able to provide the best of his assistance. This can scale up to losing thousands of lifes and crores of rupees over a year. We're aiming to build a system that allows the patient and doctor to work together as close as possible by availing all the health information of the former to the latter. Addittionally, we employ ML models to allow for predictions that can help the patient to interpret his health statistics (pure numbers) into meaningful information. A brand new doctor can easily access his the patients medical history with a simple QR scan. He can simultaneously keep track of 100s of such patients whenever he has the time. This can also reduce the needs of direct contact between the two in these testing times of COVID.

Challenges we ran into

Our biggest challenge was to get the ML model up in running as an API. We solved this to an extent. The issue we face as of now is that, some computers might not be able to access the heroku site due to some security reasons (ssh reasons). Hence we have not integrated the prediction into our app but have provided links for the github repo of our api (and all other parts) for inspection. The API link has also been provided if the judges wish to check it personally through a POST request. Also, the full API (as done to heroku ie; using docker container) is available in the master branch of the provided repo.