A DevOps team needs MLflow experiments, Unity Catalog model resources, Jobs, and Model Serving endpoints defined once and promoted through dev, staging, and production with environment-specific variables. The workload belongs to the claims triage program, uses hourly feature refreshes, and is tracked under release `mlp-106`. The exam objective is Databricks Asset Bundles. Which approach best meets the requirement?
