Training AI-300: Operationalize Machine Learning and Generative AI Solutions

The modules below are discussed during the course. For more information, view the study guide.

Design and implement an MLOps infrastructure (15–20%)

• Set up and manage Azure Machine Learning workspaces, data, and compute
• Configure security, identity and network access
• Implement infrastructure as code (Bicep, Azure CLI)
• Automate workflows with GitHub Actions

Implement machine learning model lifecycle and operations (25–30%)

• Train, evaluate, and register machine learning models
• Deploy models to production environments
• Monitor performance, detect data drift, and retrain models
• Apply responsible AI practices

Design and implement a GenAIOps infrastructure (20–25%)

• Configure Microsoft Foundry environments
• Deploy and manage generative AI models
• Manage prompts and versioning
• Secure and scale AI solutions

Implement generative AI quality assurance and observability (10–15%)

• Evaluate AI outputs using quality and safety metrics
• Monitor performance, usage, and costs
• Implement logging and debugging

Optimize AI systems and model performance (10–15%)

• Optimize RAG systems and retrieval performance
• Fine-tune models and improve accuracy
• Apply performance optimization strategies