PeopleCare Insurance Prediction October 2023

Analyzed PeopleCare’s expansion into vehicle insurance through the implementation of a predictive model for more effective customer targeting.

Technical stack Used in the Project -

• Thoroughly examined customer behavior and other relevant features through a process of data visualization and data cleaning. This ensured the availability of accurate and high-quality data for the modeling task.
• Achieved an impressive prediction accuracy of 80% by harnessing the robust capabilities of the LightGBM algorithm, optimizing its performance through hyperparameter tuning.
• Streamlined the entire model deployment process using Flask and Docker, facilitating model delivery on the Azure Container App platform, optimizing operations and enhancing scalability.

The Github code is here