Real-time Inference API and Monitoring Strategy

Develop a small client script that simulates streaming sensor data to your deployed Seldon Core endpoint and processes the anomaly predictions. Describe how you would set up monitoring for your deployed model (e.g., using Prometheus and Grafana) to track key metrics like prediction latency, model drift, and the number of detected anomalies. Outline your strategy for alerting operations personnel when a significant anomaly is detected, including the severity thresholds and communication channels.