Overview

This exercise demonstrates how to develop an AI agent using Azure AI Foundry. The agent analyzes data and generates charts dynamically using the Code Interpreter tool.

Steps & Configuration Details

1. Create an Azure AI Foundry Project

  • Open Azure AI Foundry portal (https://ai.azure.com) and sign in.
  • Select Create an agent.
  • Configuration Items:
    • Azure AI Foundry Resource: A valid name.
    • Subscription: Your Azure subscription.
    • Resource Group: Select or create a resource group.
    • Region: Choose any AI Services-supported location.
    • Project Name: A unique name.

After creation, the Agents playground opens automatically, with a GPT-4o base model deployed.

2. Configure the Agent Client App

  • Open Azure Portal (https://portal.azure.com).
  • Launch Azure Cloud Shell (PowerShell environment).
  • Clone the repository:
    rm -r ai-agents -f
git clone https://github.com/MicrosoftLearning/mslearn-ai-agents ai-agents
  • Navigate to the correct folder:
    cd ai-agents/Labfiles/02-build-ai-agent/Python
  • Install dependencies:
    python -m venv labenv
./labenv/bin/Activate.ps1
pip install -r requirements.txt azure-ai-projects
  • Open the configuration file:
    code .env
  • Update Configuration Values:
    • Project Endpoint (copied from Azure AI Foundry portal).
  • Save the configuration file.

3. Write Code for the Agent App

  • Open the code file:
    code agent.py
  • Add references:
    from azure.identity import DefaultAzureCredential
from azure.ai.agents import AgentsClient
from azure.ai.agents.models import FilePurpose, CodeInterpreterTool, ListSortOrder, MessageRole
  • Connect to the Agent client:
    agent_client = AgentsClient(
    endpoint=project_endpoint,
    credential=DefaultAzureCredential(
        exclude_environment_credential=True,
        exclude_managed_identity_credential=True
    )
)
with agent_client:
  • Upload the data file and create a CodeInterpreterTool:
    file = agent_client.files.upload_and_poll(
    file_path=file_path,
    purpose=FilePurpose.AGENTS
)
print(f"Uploaded {file.filename}")

code_interpreter = CodeInterpreterTool(file_ids=[file.id])
  • Define an agent that uses the CodeInterpreterTool:
    agent = agent_client.create_agent(
    model=model_deployment,
    name="data-agent",
    instructions="You are an AI agent that analyzes the data in the file that has been uploaded. If the user requests a chart, create it and save it as a .png file.",
    tools=code_interpreter.definitions,
    tool_resources=code_interpreter.resources,
)
print(f"Using agent: {agent.name}")
  • Create a thread for the conversation:
    thread = agent_client.threads.create()
  • Send a prompt to the agent:
    message = agent_client.messages.create(
    thread_id=thread.id,
    role="user",
    content=user_prompt,
)
run = agent_client.runs.create_and_process(thread_id=thread.id, agent_id=agent.id)

if run.status == "failed":
    print(f"Run failed: {run.last_error}")
  • Show the latest response from the agent:
    last_msg = agent_client.messages.get_last_message_text_by_role(
    thread_id=thread.id,
    role=MessageRole.AGENT,
)
if last_msg:
    print(f"Last Message: {last_msg.text.value}")
  • Get the conversation history:
    print("\nConversation Log:\n")
messages = agent_client.messages.list(thread_id=thread.id, order=ListSortOrder.ASCENDING)
for message in messages:
    if message.text_messages:
        last_msg = message.text_messages[-1]
        print(f"{message.role}: {last_msg.text.value}\n")
  • Retrieve generated files:
    for msg in messages:
    for img in msg.image_contents:
        file_id = img.image_file.file_id
        file_name = f"{file_id}_image_file.png"
        agent_client.files.save(file_id=file_id, file_name=file_name)
        print(f"Saved image file to: {Path.cwd() / file_name}")
  • Clean up:
    agent_client.delete_agent(agent.id)

4. Run the Agent App

  • Sign into Azure:
    az login
  • Run the application:
    python agent.py
  • Example prompt:
    What's the category with the highest cost?
  • Request a chart:
    Create a pie chart showing cost by category.
  • The agent should generate a .png file with the requested chart.

5. Clean Up

  • Delete Azure resources to avoid unnecessary costs:

Related Posts