MLOps in Production — Monitoring & Autoscaling

Monitor live model performance, detect data drift and concept drift, and set up automated autoscaling.

Assumes: Observability & Monitoring and Kubernetes-Native Model Serving.

Modules

1. Introduction to ML Observability (Data Drift vs. Concept Drift)
2. Metrics, Logging, and Alerting for Deployed Models
3. Horizontal Pod Autoscaling (HPA) and Resource Optimization

Tools

• Prometheus
• Grafana
• Evidently AI
• Kubernetes HPA
• Linux