Open-source Python library for evaluating ML model reliability beyond accuracy — with calibration, failure, and fairness diagnostics for informed deployment decisions.
