nannyml

Testing out NannyML

Collibra governance tooling — edge stack deployment, contract ID generation, data contract lifecycle, ORM/SBVR semantic pipeline, and API server. All features are exposed through the singine collibra CLI family.

Training-serving feature skew detection with LightGBM, MLflow, BentoML, Prometheus, Grafana, and NannyML.

Production-oriented MLOps workflow for NYC Green Taxi tip prediction using Metaflow, MLflow, data-quality gates, drift monitoring, automated retraining, and governed model promotion.

This repository describes how to monitor a model using NannyML

ML model monitoring for fraud detection using NannyML

Help the bank monitoring their fraud detection model and figuring out why it's not performing as expected.

On-call AI agent that detects ML model drift, walks DataHub lineage to the upstream root cause, identifies the owner, and writes a drift-causation object back onto the model.