MLOps Engineer — Model Monitoring & Drift Detection

We have 8 models in production. None of them are monitored. We find out a model is degrading when a business stakeholder complains. That's too late. We need monitoring set up: data drift alerts, prediction distribution shifts, business metric correlation, and automated retraining triggers when thresholds are breached. We're open to Evidently AI, WhyLabs, or a custom solution — whatever you think is right for our scale. We run on GCP. Models are served via FastAPI on Cloud Run. This is 3 months of setup + documentation. After that, our team takes over maintenance.