Collection Management is an AI-powered worklist that prioritizes the top 20% of delinquent customers, automating collections for the remaining 80% of long-tail customers.

As an intern at Highradius, I collaborated with the ML team to develop and deploy predictive models with 60-80% accuracy, improving the client's collection management efficiency.

My key responsibilities included:

• Extracting and preprocessing data from the client's database to prepare it for model training.
• Developing and deploying multiple predictive models in a production environment to streamline the prediction process.
• Documenting the entire deployment process, including UI navigation for seamless integration, model training steps, and operational guidelines.

• StackPython, Pandas, Numpy, Scikit-learn, Tensorflow, Keras
• InfoCollections Management Software