Skip to content

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:

az login
az account show

Python

Install Python 3.10 or later. Verify with:

python --version

Create a Microsoft Foundry project

  1. Go to the Azure portal and create an Microsoft Foundry resource (or use an existing one).
  2. Inside the resource, create a project.
  3. 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.

  1. Open your project in Microsoft Foundry.
  2. Select Build in the top navigation, then Models in the left pane.
  3. Select Deploy a base model, choose gpt-5.4-mini, and complete the deployment wizard.
  4. Note the deployment name — you will use it as FOUNDRY_MODEL in your .env file.

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

cp .env.example .env

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.

  1. Open the Extensions panel (Ctrl+Shift+X).
  2. Search for AI Toolkit (ms-windows-ai-studio.windows-ai-studio).
  3. Click Install.

Verify your setup

Run a quick check to make sure everything works:

az account show          # should show your subscription
python -c "import agent_framework; print('OK')"

If both commands succeed, you are ready to start Lesson 01.

Official references