Solid knowledge of end-to-end ML systems, including training data, features, model artifacts, offline validation, online serving, post-deployment metrics, and business outcome measurement. Within the first 90 days, you will have assessed the current production ML operational landscape, identified the highest-risk gaps, and established a prioritized roadmap for observability, traceability, deployment safety, and incident response.