Role Mandate: Build the technical controls, automation, and deployment pipelines that make the centralized MLOps operating model work in practice.
Key Responsibilities
Build CI/CD pipelines for controlled promotion from development/test to production using Azure DevOps or GitHub Actions.
Implement MLflow tracking and Unity Catalog-based model registry and promotion patterns.
Create reusable Databricks Jobs and Workflows for training, validation, registration, deployment, and retraining.
Automate setting provisioning and controls using Terraform and Infrastructure as Code.
Implement RBAC, service principals, managed identities, Key Vault integration, and production access restrictions.
Build model serving, batch inference, monitoring, alerting, and rollback patterns.
Implement usage logging, tags, compute policies, and cost telemetry for market and use-case attribution.
Build guardrails so market users can work independently in sandbox and development settings without direct production deployment rights.
Operationalize RAG and GenAI workloads with evaluation, observability, security, and cost controls.