⚓ Eurybia monitors model drift over time and securizes model deployment with data validation

A curated list of awesome open source tools and commercial products for monitoring data quality, monitoring model performance, and profiling data 🚀

In this repository, we will present techniques to detect covariate drift, and demonstrate how to incorporate your own custom drift detection algorithms and visualizations with SageMaker model monitor.

These are my notes of the Udacity Nanodegree Machine Learning DevOps Engineer.

Simulation, testing and comparison of state of the art Unsupervised Concept Drift Detectors used in a batch Machine Learning scenario.

資料科學的日常研究議題

Detect behavioural drift between LLM versions before you upgrade. Compare model responses, classify regressions, and generate migration reports with validated prompt patches.

In this project, we illustrate how the Kolmogorov Smirnov (KS) statistical test works, and why it is commonly used in Machine Learning (ML), Deep Learning (DL) and Artificial Intelligence (AI).

Learn how to handle model drift and perform test-based model monitoring

From model.fit() to models that survive production — a phased climb through ML fundamentals and the MLOps layer: versioning, serving, drift, scale.

Time‑aware NBA forecasting pipeline (R² 0.94 points) with rolling CV, leakage guards, and automated retraining; includes backtesting reports and model card.

Production-style ML monitoring template on the Wine Quality (red) dataset: Evidently (data/target/prediction drift, data quality) + adversarial validation, PSI/JS effect sizes, SHAP/PDP, slice analysis, and an Alert Policy with actions

ModelPulse helps maintain model reliability and performance by providing early warning signals for these issues, allowing teams to address them before they impact users significantly.

Decision DNA is an AI governance and monitoring platform designed to supervise machine learning models used in credit risk decision systems. The system helps detect model drift, operational risks, and security threats, while maintaining transparent and auditable AI decision pipelines.

"Past performance of machine learning model is no guarantee of future results." We call it "model drift" or "model decay". This repository will introduce various methods for detecting model drift.

Diagnostic test suite for measuring whether AI models preserve the named Origin boundary inside AI Foundations / Origin | Continuum.

Tracks ML model drift and data quality in real-time

LLMs as production extraction infrastructure: rule-vs-LLM triage, validated structured outputs, cost-capped model routing, a measured precision/recall eval + vendor-drift detection, entity resolution, and an Airflow 3 ETL DAG. Runs fully offline.

Quantifying the "Safety Half-Life" of LLMs: A framework to measure how safety alignment degrades and susceptibility to jailbreaks increases as context length grows

An ML monitoring framework, applied to an attrition risk assessment system.