Context

A system designed for healthcare, where AI must support — not replace — the decisions of those who care for patients.

The challenge

Heterogeneous, sensitive data and, above all, the need for interpretable models: in healthcare a black-box prediction is not enough.

The solution

A data-preparation pipeline and predictive models chosen by balancing accuracy and interpretability, with readable output for decision support.

The result

A tool to support clinical decisions, designed to be reliable and understandable.

Stack: Python · Machine Learning · Healthcare