DS305: Model Deployment & MLOps

by

SFIA Reference: INCA / SWDN Level 4/5

Data Scientist (Grade 3)

Competency Description

Packages and deploys models into production pipelines and maintains model lifecycle using MLOps practices.

Learning Outcomes

  • Deploy models using MLflow, Azure ML, or FastAPI.
  • Version control model code and configurations.
  • Monitor model performance and data drift.

Evidence Requirements

Level 4:

  • Operationalised a model for batch or real-time use.
  • Used Git and CI/CD pipelines for deployment. Level 5:
  • Built a retraining pipeline or monitoring framework.
  • Provided guidance on model lifecycle management.

Suggested Learning Activities

  • Deploy a simple model to Azure ML with MLflow.
  • CI/CD lab: GitHub Actions for model updates.
  • Case study: Response to model performance decay.

Recommended Resources

  • Microsoft Learn: “MLOps with Azure Machine Learning”
  • Internal repo: Model Deployment Templates

Leave a Comment