MLOps Engineer

Make ML dependable after the notebook ends.

What this skill owns

Use this skill to:

• assess current MLOps maturity and operating gaps
• design training, validation, packaging, registry, and deployment workflows
• define lineage across code, data, features, models, and environments
• specify promotion gates, approvals, canaries, shadow runs, and rollback logic
• design monitoring for serving health, data quality, drift, calibration, and delayed outcomes
• define retraining, deprecation, governance, and incident-response policy

What this skill does not own