00 — Prerequisites
Before you start the workshop, make sure the following items are in place.
Azure subscription
You need an active Azure subscription. If you do not have one, create a free account.
Azure CLI
Install the Azure CLI (version 2.67.0 or later) and sign in:
Python
Install Python 3.10 or later. Verify with:
Create a Microsoft Foundry project
- Go to the Azure portal and create an Microsoft Foundry resource (or use an existing one).
- Inside the resource, create a project.
- Copy the project endpoint — it looks like:
https://<your-resource>.services.ai.azure.com/api/projects/<your-project>
Tip
The project endpoint is shown on the project overview page in the Azure portal.
Deploy a model
Inside your Foundry project, deploy a model. This workshop recommends gpt-5.4-mini because it is fast and cost-effective for learning.
- Open your project in Microsoft Foundry.
- Select Build in the top navigation, then Models in the left pane.
- Select Deploy a base model, choose
gpt-5.4-mini, and complete the deployment wizard. - Note the deployment name — you will use it as
FOUNDRY_MODELin your.envfile.
Install workshop dependencies
python -m venv .venv
source .venv/bin/activate # macOS / Linux
# .venv\Scripts\activate # Windows
pip install --upgrade pip
pip install -r requirements.txt
Configure environment variables
Open .env and set the two required values:
FOUNDRY_PROJECT_ENDPOINT=https://<your-resource>.services.ai.azure.com/api/projects/<your-project>
FOUNDRY_MODEL=gpt-5.4-mini
VS Code extensions (optional)
For Lesson 06 (Observability), install the AI Toolkit extension in VS Code. It provides a visual trace viewer for OpenTelemetry spans emitted by your agents.
- Open the Extensions panel (
Ctrl+Shift+X). - Search for AI Toolkit (ms-windows-ai-studio.windows-ai-studio).
- Click Install.
Verify your setup
Run a quick check to make sure everything works:
If both commands succeed, you are ready to start Lesson 01.