Microsoft Foundry Quickstart

Get started with Microsoft Foundry SDK building AI applications with agents, models, and tools.

Microsoft Foundry Quickstart

This quickstart will help you get started with Microsoft Foundry by creating a project, deploying a model, running chat completions, and creating an agent with file search capabilities.

Prerequisites

• An Azure subscription (Create one for free)
• Azure CLI installed and authenticated
• Python 3.8 or later (for Python examples)

Step 1: Create a Foundry Project

A Foundry project is where you do most of your development work. Projects provide isolation for agents, models, and data.

1. Sign in to Azure

   az login
   
2. Create a Foundry resource and project

   You can create resources using the Azure portal or deploy using Azure CLI with Bicep templates.

   For a basic setup:

   # Deploy basic agent setup
   az deployment group create \
     --resource-group <your-resource-group> \
     --template-uri https://raw.githubusercontent.com/azure-ai-foundry/foundry-samples/main/infrastructure/infrastructure-setup-bicep/40-basic-agent-setup/azuredeploy.json
   

   This creates:

   • Foundry account
   • Foundry project
   • GPT-4.1 model deployment
3. Get your project endpoint

   After deployment, navigate to your project in the Azure portal and copy the endpoint URL:

   https://<resource-name>.services.ai.azure.com/api/projects/<project-name>
   

Step 2: Install the SDK

Python

pip install openai azure-identity azure-ai-projects==1.0.0

Step 3: Chat with a Model

Create your first chat completion using a deployed model.

Python

from azure.ai.projects import AIProjectClient
from azure.identity import DefaultAzureCredential

# Initialize project client
project = AIProjectClient(
    endpoint="https://<resource-name>.services.ai.azure.com/api/projects/<project-name>",
    credential=DefaultAzureCredential(),
)

# Get OpenAI client
models = project.get_openai_client(api_version="2024-10-21")

# Create chat completion
response = models.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You are a helpful assistant"},
        {"role": "user", "content": "What is the size of France in square miles?"},
    ],
)

print(response.choices[0].message.content)

Step 4: Create an Agent

Create an intelligent agent that can assist with tasks.

Python

from azure.ai.projects import AIProjectClient
from azure.identity import DefaultAzureCredential
import os
import time

# Initialize client
project = AIProjectClient(
    endpoint=os.environ["PROJECT_ENDPOINT"],
    credential=DefaultAzureCredential(),
)

# Create agent
agent = project.agents.create_agent(
    model=os.environ["MODEL_DEPLOYMENT_NAME"],
    name="my-agent",
    instructions="You are a helpful assistant",
)
print(f"Created agent, ID: {agent.id}")

# Create thread
thread = project.agents.threads.create()
print(f"Created thread, ID: {thread.id}")

# Send message
message = project.agents.messages.create(
    thread_id=thread.id,
    role="user",
    content="Hello, what can you help me with?",
)

# Run agent
run = project.agents.runs.create(thread_id=thread.id, agent_id=agent.id)

# Wait for completion
while run.status in ["queued", "in_progress"]:
    time.sleep(1)
    run = project.agents.runs.get(thread_id=thread.id, run_id=run.id)

print(f"Run completed with status: {run.status}")

# Get messages
messages = project.agents.messages.list(thread_id=thread.id)
for msg in messages:
    print(f"{msg['role']}: {msg['content']}")

# Cleanup
project.agents.delete_agent(agent.id)

Step 5: Add File Search to Your Agent

Enhance your agent with knowledge retrieval capabilities.

Python

from azure.ai.projects import AIProjectClient
from azure.ai.projects.models import FileSearchTool, FilePurpose
from azure.identity import DefaultAzureCredential
import os

project = AIProjectClient(
    endpoint=os.environ["PROJECT_ENDPOINT"],
    credential=DefaultAzureCredential(),
)

# Upload file
file = project.agents.files.upload_and_poll(
    file_path="product_info_1.md",
    purpose=FilePurpose.AGENTS
)
print(f"Uploaded file, file ID: {file.id}")

# Create file search tool with file
file_search_tool = FileSearchTool(file_ids=[file.id])

# Create agent with file search
agent = project.agents.create_agent(
    model=os.environ["MODEL_DEPLOYMENT_NAME"],
    name="file-search-agent",
    instructions="You are a helpful assistant that can search information from uploaded files.",
    tools=file_search_tool.definitions,
    tool_resources=file_search_tool.resources,
)

print(f"Created agent with file search, agent ID: {agent.id}")

# Create thread and ask question
thread = project.agents.threads.create()
message = project.agents.messages.create(
    thread_id=thread.id,
    role="user",
    content="What products does Contoso offer?",
)

# Run and wait for completion
run = project.agents.runs.create_and_poll(
    thread_id=thread.id,
    agent_id=agent.id
)

if run.status == "completed":
    messages = project.agents.messages.list(thread_id=thread.id)
    for msg in messages:
        print(f"{msg['role']}: {msg['content']}")

# Cleanup
project.agents.delete_agent(agent.id)
project.agents.files.delete(file.id)

Next Steps

Now that you’ve created your first agent with file search, explore more capabilities:

Agent Overview

Learn about Foundry Agent Service capabilities

Environment Setup

Configure your agent environment

Agent Tools

Explore built-in tools for agents

SDK Overview

Deep dive into the Foundry SDK

Clean Up Resources

To avoid incurring charges, delete the resources you created:

az group delete --name <your-resource-group> --yes --no-wait