Environment Setup for Azure AI Foundry Workshop¶

This notebook will guide you through setting up your environment for the Azure AI Foundry workshop.

Prerequisites¶

Python 3.8 or later
Azure subscription with AI services access
Basic Python knowledge

In [ ]:

  Copied!     
 
# Install required packages (if not already installed):
!pip install azure-identity azure-ai-projects
# Install required packages (if not already installed): !pip install azure-identity azure-ai-projects

Azure Authentication Setup¶

First, we'll verify our Azure credentials and setup.

In [ ]:

  Copied!     
 
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient
from azure.ai.projects.models import ConnectionType
import os

# Initialize Azure credentials
try:
    credential = DefaultAzureCredential()
    print("✓ Successfully initialized DefaultAzureCredential")
except Exception as e:
    print(f"× Error initializing credentials: {str(e)}")
from azure.identity import DefaultAzureCredential from azure.ai.projects import AIProjectClient from azure.ai.projects.models import ConnectionType import os # Initialize Azure credentials try: credential = DefaultAzureCredential() print("✓ Successfully initialized DefaultAzureCredential") except Exception as e: print(f"× Error initializing credentials: {str(e)}")

Initialize AI Project Client¶

Note: Before proceeding, ensure you:

Copy your .env.local file to .env

Update the project connection string in your .env file

Have a Hub and Project already provisioned in Azure AI Foundry

You can find your project connection string in Azure AI Foundry under your project's settings:

Understanding AIProjectClient¶

The AIProjectClient is a key component for interacting with Azure AI services that:

Manages Connections: Lists and accesses Azure AI resources like OpenAI models
Handles Authentication: Securely connects using Azure credentials
Enables Model Access: Provides interfaces to use AI models and deployments
Manages Project Settings: Controls configurations for your Azure AI project

The client requires:

A project connection string (from Azure AI project settings)
Valid Azure credentials

You can find your project connection string in Azure AI Studio under Project Settings:

In [ ]:

  Copied!     
 
from dotenv import load_dotenv

# Create AI Project client
# Load environment variables from .env file
load_dotenv()

# Initialize AIProjectClient with connection string and credentials
try:
    client = AIProjectClient.from_connection_string(
        conn_str=os.getenv("PROJECT_CONNECTION_STRING"),
        credential=credential
    )
    print("✓ Successfully initialized AIProjectClient")
except Exception as e:
    print(f"× Error initializing client: {str(e)}")
from dotenv import load_dotenv # Create AI Project client # Load environment variables from .env file load_dotenv() # Initialize AIProjectClient with connection string and credentials try: client = AIProjectClient.from_connection_string( conn_str=os.getenv("PROJECT_CONNECTION_STRING"), credential=credential ) print("✓ Successfully initialized AIProjectClient") except Exception as e: print(f"× Error initializing client: {str(e)}")

Verify Access to Models¶

Finally, let's verify we can access the available models.

In [ ]:

  Copied!     
 
# List the properties of all connections
connections = client.connections.list()
print(f"====> Listing of all connections (found {len(connections)}):")
for connection in connections:
    print(connection)

# List the properties of all connections of a particular "type" (in this sample, Azure OpenAI connections)
connections = client.connections.list(
    connection_type=ConnectionType.AZURE_OPEN_AI,
)
print(f"====> Listing of all Azure Open AI connections (found {len(connections)}):")
for connection in connections:
    print(connection)

# Get the properties of the default connection of a particular "type", with credentials
connection = client.connections.get_default(
    connection_type=ConnectionType.AZURE_AI_SERVICES,
    include_credentials=True,  # Optional. Defaults to "False"
)
print("====> Get default Azure AI Services connection:")
print(connection)

print("====> Get connection by name:")
print(connection)
# List the properties of all connections connections = client.connections.list() print(f"====> Listing of all connections (found {len(connections)}):") for connection in connections: print(connection) # List the properties of all connections of a particular "type" (in this sample, Azure OpenAI connections) connections = client.connections.list( connection_type=ConnectionType.AZURE_OPEN_AI, ) print(f"====> Listing of all Azure Open AI connections (found {len(connections)}):") for connection in connections: print(connection) # Get the properties of the default connection of a particular "type", with credentials connection = client.connections.get_default( connection_type=ConnectionType.AZURE_AI_SERVICES, include_credentials=True, # Optional. Defaults to "False" ) print("====> Get default Azure AI Services connection:") print(connection) print("====> Get connection by name:") print(connection)

Validate Model and Search Connections¶

The cell below validates that we have properly provisioned and connected to:

Azure OpenAI models through our Azure OpenAI connection
Azure AI Search through our Azure AI Search connection

Both of these services will be essential for building our AI applications. The OpenAI models will provide the core language capabilities, while Azure AI Search will enable efficient information retrieval and knowledge base functionality.

In [ ]:

  Copied!     
 
# List all connections and check for specific types
conn_list = client.connections.list()
search_conn_id = ""
openai_conn_id = ""

for conn in conn_list:
    conn_type = str(conn.connection_type).split('.')[-1]  # Get the part after the dot
    if conn_type == "AZURE_AI_SEARCH":
        search_conn_id = conn.id
    elif conn_type == "AZURE_OPEN_AI":
        openai_conn_id = conn.id

print(f"\n====> Connection IDs found:")
if not search_conn_id:
    print("Azure AI Search: Not found - Please create an Azure AI Search connection")
else:
    print(f"Azure AI Search: {search_conn_id}")
    
if not openai_conn_id:
    print("Azure OpenAI: Not found - Please create an Azure OpenAI connection") 
else:
    print(f"Azure OpenAI: {openai_conn_id}")
# List all connections and check for specific types conn_list = client.connections.list() search_conn_id = "" openai_conn_id = "" for conn in conn_list: conn_type = str(conn.connection_type).split('.')[-1] # Get the part after the dot if conn_type == "AZURE_AI_SEARCH": search_conn_id = conn.id elif conn_type == "AZURE_OPEN_AI": openai_conn_id = conn.id print(f"\n====> Connection IDs found:") if not search_conn_id: print("Azure AI Search: Not found - Please create an Azure AI Search connection") else: print(f"Azure AI Search: {search_conn_id}") if not openai_conn_id: print("Azure OpenAI: Not found - Please create an Azure OpenAI connection") else: print(f"Azure OpenAI: {openai_conn_id}")