Baby, let's move!

Key features
dashboard to track your exercise routine
chatbot to answer all your queries
blog page to stay updated

Inspiration

What a bizarre world we’re currently living in! For women who are pregnant, this can be a particularly emotional and stressful time. Taking some time for one’s own mental and physical health is very important and for most, this should include exercising.

With current restrictions, most of us don’t have access to gyms or professional trainers so we resort to exercise by following online content (fitness apps, youtube videos, etc).

Now, the problem with this is that exercise, if not done correctly can do more harm than good. Incorrect posture could cause joint pain, back pain, and propose some serious health concerns for the mother and the baby.

Keeping all these points in mind, we came up with an app for pregnant women where they can follow an exercise regime: as you follow the exercise, the app keeps a check on your posture. If at any point your posture is incorrect, the app instructs you on how to correct it. All this happens in real-time, as you exercise.

Apart from this, we added two more key features which include: A chatbot backed up by AI which answers your health-related question And a blogs section where you can find certified articles about exercise and lifestyle during pregnancy.

Features

An app where pregnant women can follow an exercise regime. With the help of Machine Learning posenet model, the app helps pregnant women to complete different exercises and keeps a check on the user’s posture during exercise, and informs and instructs them whenever their posture is not proper.
An AI chatbot based on NLP and Google's DialogFlow where the user can ask all their health and exercise-related questions related to pregnancy. The chatbot is available 24/7 to assist users with their queries.
A blogs section where the user can find certified articles based on pregnant women’s healthcare.

How we built it

Android app: We used Flutter to create a cross-platform app and Google's Cloud Firestore to store users' databases like activity logs and chat database.
Exercise Model: We used Tensorflow's Posenet model to track the user's poses while they exercise and simultaneously guide them to improve their posture.
Chatbot: We used Natural Language Processing for the chatbot. With the help of sentiment analysis, the chatbot predicts the user's mood and even gives answers based on the same.