CostCareAI

An AI-powered web application that uses user information, such as age, BMI, and region etc, to forecast annual healthcare costs can help users make better plans and feel less stressed about money.

Comment

Inspiration

People frequently don't know how much to budget or whether they have too much or too little insurance because healthcare costs can be stressful and unpredictable. My goal was to use AI to make personal healthcare finances more transparent and manageable.

What it does

A web application called CostCareAI forecasts your annual medical costs by taking into account important lifestyle and individual characteristics such as age, body mass index, smoking status, and more. After users enter their information, the app instantly generates a cost estimate driven by AI and displays insights that highlight the factors that influence those costs.

How we built it

To examine past healthcare cost data, I used Python (pandas, NumPy, scikit-learn) to train a regression model. For real-time predictions, the model was incorporated into an interactive Streamlit web application. Power BI, Excel, and SQL were used for data processing, cleaning, and visualisation, guaranteeing accuracy and interpretability.

Challenges we ran into

locating and sanitising machine learning-ready healthcare cost datasets.

maximising accuracy while avoiding overfitting in the regression model.

maintaining the model's usability while smoothly integrating it into a Streamlit interface.

Accomplishments that we're proud of

created a functional AI model that generates accurate forecasts of healthcare expenses.

created a simple, user-friendly interface to make it easy for non-technical users to examine the results.

developed a pipeline that runs from data preprocessing to deployment.

What we learned

Practical implementation of regression models in real-world healthcare use cases.

Deploying ML models using Streamlit.

How transparency in AI predictions can improve user trust and decision-making.

What's next for CostCareAI

extending to international datasets in order to increase the coverage of predictions.

incorporating additional risk factors (such as medication use and chronic conditions).

incorporating safe APIs to compare insurance quotes in real time.

“This project was built as part of LIVE AI Best Coast 2025 under the Global (online) participation track.”

Built With

Share this project:

Updates