PregCare

PregCare

Empowering Moms-to-be everywhere ; your virtual support system for a healthy pregnancy journey

The problem PregCare solves

Many pregnant women find themselves navigating the challenges of pregnancy without the support of their families due to distance or other circumstances, resulting in a lack of access to essential information and guidance. This absence of familial support can lead to feelings of isolation and uncertainty, exacerbating concerns about maternal health and well-being.
In response, there is a pressing need for a comprehensive online platform that offers tailored support, expert advice, and resources to empower pregnant women who are away from their families to navigate their pregnancy journey with confidence and ease.
PregCare offers features like predictive analysis , tracker , nutritional information guide , curated yoga , breastfeeding guide , blogs of support and a personal chatbot to resolve such problems.

Challenges we ran into

Bug- incorporating an authentication system as first year students
Fixes- Used and learned implementing Auth0 to make it easier
Bug- while using Auth0 it threw an error of mismatch of callbackURL type asking for a secured https link , which we could not due to the local host issues .
Fixes-upon debugging the whole codebase again we witnessed an error in the window.origin.location for the redirect call and changed it to the required index.html url
Bug-while training the ML model we tried to implemented KNN also but it threw some error which we were unable to debug for the majority of time and due to time constraints we had to move to the frontend part
Fixes- upon seeking help from online resources , chatgpt, kaggle we were able to fix it but as the results were the best for the random forest model we did not include it ad we already had the better model
Bug-connecting ML model files (.ipynb & .py) with the javascript frontend was very tough and I could not really find a solution for it
Fixes-Debugged it for long time but then we decided to host the ML model over streamlit and connect it as a URL , learnt streamlit as something new
another problem faced was that I was the only person coding in the team so with the time limits and classes it was very hard to manage , fixes include skipping sleep :)

Added the filters to AI chatbot now it wont return java codes , added responsiveness for various pages and Logout button

Discussion