NannyML is an open-source Python library focused on post-deployment data science and ML model monitoring. It detects silent model failures by estimating model performance when ground truth is unavailable and identifying data drift in production. The library provides visualization and alerting capabilities to help data scientists maintain model quality over time.