Diabeater

Inspiration

We were interested in using health data and we created a very impressive model. We decided to focus or project around it as a result

What it does

It uses a predictive AI model to tell a user there chances of developing diabetes

How we built it

The model was built using python tensorflow, keras and sckit-learn. We used Flask to serve the model and express as the backend api for our next js application

Challenges we ran into

We found difficulties integrating certain APIs and ended up dropping certain technologies instead focusing on the model. we found connecting the various servers to be a challenge

Accomplishments that we're proud of

We where able to create an extremely accurate predictive model that does not require massive amounts of data

What we learned

We learned how to use TensorFlow keras alongside becoming more proficient in javascript, next js and flask.

What's next for Diabeater

We aim to add more features including user authentication and give users the ability to plan out how to reduce there chances.

Built With

• express.js
• flask
• keras
• nextjs
• tensorflow