What this course is
A hands-on course for the AI-300 exam.
Act 1 takes a model through the full machine learning lifecycle in Azure Machine Learning, from workspace to CI/CD and security.
Act 2 runs generative AI in Microsoft Foundry: model deployment, agents, RAG, fine-tuning, evaluation and security. Most theory lectures are followed by a lab you run in your own Azure subscription.
What you get
- The full course: 13 hours of video.
- Course visuals: every lecture graphic as 4K slides in one PDF, plus the diagrams in a zoomable map viewer.
- Source code: the lab code from the course.
- Practice questions, taken in the browser.
- A study guide, as a PDF.
Who this is for
- Engineers preparing for the AI-300 exam (Microsoft Certified: Machine Learning Operations Engineer Associate)
- Machine learning and MLOps engineers who build training and deployment pipelines on Azure
- AI and platform engineers who run generative AI applications in Microsoft Foundry
- Data scientists who want to take their models into production on Azure
What you need
- An Azure subscription where you are Owner, or Contributor plus User Access Administrator. Azure usage is billed to that subscription, separately from the course.
- Basic Azure knowledge: subscriptions, resource groups, and the Azure portal.
- Python 3.10 to 3.12, Visual Studio Code with the Python and Jupyter extensions, the Azure CLI, and a bash shell.
- Basic Python: reading and running scripts and notebooks.
- A GitHub account and the GitHub CLI for the CI/CD for Machine Learning labs.
Course curriculum
- Lab Setup
- Machine Learning (18:37)
- Azure Machine Learning (13:03)
- Microsoft Foundry vs. Azure Machine Learning (11:41)
- MLOps (17:20)
- What We're Building (2:51)
- Create a Workspace and Tour Azure ML Studio (14:56)
- Configuring Datastores and Compute (15:05)
- Practice quiz · MLOps Infrastructure and Workspace Management
- Answers explained · MLOps Infrastructure and Workspace Management
- Training Jobs, Hyperparameter Tuning, and Pipelines (13:47)
- MLflow Tracking and Logging (8:13)
- Training Scripts and Command Jobs (10:57)
- Running Sweep Jobs (14:27)
- Running AutoML Jobs (13:12)
- Building a Training Pipeline (15:59)
- Practice quiz · Orchestrating Model Training
- Answers explained · Orchestrating Model Training
- Microsoft Foundry Overview (14:07)
- What is GenAIOps? (12:41)
- Foundry Resource and Project Architecture (14:29)
- Foundry Pricing (4:27)
- Create a Foundry Project and Tour the Portal (19:36)
- Deploy Foundry (14:11)
- Practice quiz · Foundry Platform and GenAIOps Fundamentals
- Answers explained · Foundry Platform and GenAIOps Fundamentals
- Foundry Models (9:04)
- Model Deployment Options: Serverless, Managed Compute, and PTUs (11:29)
- Deploy an OpenAI model with the Foundry portal (14:28)
- Deploy an OpenAI model with Python (12:21)
- Chat with an OpenAI model using Python (2:20)
- Practice quiz · Deploying and Managing Foundation Models
- Answers explained · Deploying and Managing Foundation Models
- Foundry Agent Service (3:06)
- Create an agent with the Foundry portal (11:50)
- Create your first agent with Python (13:23)
- Create a multi tool agent with Python (7:52)
- Build an Agent and Improve It with the Agent Optimizer (16:21)
- Practice quiz · Foundry Agent Service
- Answers explained · Foundry Agent Service
- Common AI risks (7:52)
- Prompt Injection (5:32)
- Training Data Poisoning (4:56)
- Overreliance (3:56)
- Foundry Identity, Networking, and Content Safety (12:51)
- Microsoft Defender for Cloud (MDC) (15:24)
- Microsoft Defender for Cloud - Defender for AI Services (4:18)
- Managed Identity and RBAC Hardening (8:33)
- Microsoft Defender for Cloud - Configure Defender for AI Services (8:03)
- Practice quiz · AI Risks and Foundry Security
- Answers explained · AI Risks and Foundry Security
About the instructor
Christopher Nett
Security Architect @ Microsoft · Content Creator
I have worked in cybersecurity since 2014, from incident response to cloud security to security architecture, and joined Microsoft in 2022. Since 2023 I have published courses on Microsoft Security, Azure and AI.
- CISSP
- CCSP
- CISM
- M.Sc. IT Security
- MBA
Frequently Asked Questions
Can I download the videos?
No. The videos stream on the platform. The source code, the course visuals and the study guide are downloads.
How long do I have access?
It is a one-time purchase. Access and use are set out in the Terms of Sale.
What if the course is not right for me?
You have a 14-day money-back guarantee.
Is there a community?
Yes. Join the Discord community: https://discord.gg/QfC2RJ2ZYN. It is free and open to everyone.