ML Drift Monitoring in Production: What We Measure and Why

Feature distribution drift is not the same as label drift and neither is the same as business metric degradation. Confusing them leads to alert fatigue.

Overview

This note is part of the field-notes archive generated for this site. The summary below is the published excerpt; you can expand the full write-up anytime in the CMS.

Series

Part of ML in Production (installment 1).

Related notes

• Quantization War Stories Llm
• Ml Paper Reproducibility Notes

Tags

• machine-learning
• mlops
• monitoring
• drift
• production