02 Exercise 🛠️: Build an AI Agent with Code Interpreter

In this hands-on lab, you will use Azure AI Agent Service to build an agent that analyzes data and dynamically generates charts using the built-in Code Interpreter tool.

📌 Table of Contents

1. Create Azure AI Foundry Project
2. Create an Agent Client App
3. Configure Application Settings
4. Write Code for the Agent App
5. Sign into Azure and Run the App
6. 📊 Summary
7. 🧹 Clean Up

---

🔧 Create Azure AI Foundry Project

1. Open https://ai.azure.com and sign in.
2. Click Create an agent.
3. When prompted:
4. Provide a valid Project name
5. Expand Advanced options:
   • Azure AI Foundry resource: your resource name
   • Subscription: your Azure subscription
   • Resource group: existing or new
   • Region: any supported region
6. Click Create and wait for provisioning.
7. The Agents Playground will open automatically.
8. In the left pane, click Overview.
9. Copy your project endpoint and save it for later use.

---

💻 Create an Agent Client App

1. Clone the provided GitHub repo:

git clone https://github.com/MicrosoftLearning/mslearn-ai-agents ai-agents
cd ai-agents/Labfiles/02-build-ai-agent/Python
ls -a -l

2. The folder includes: - Application code - Configuration file (.env) - Sample data file: data.txt

⚙️ Configure Application Settings

1. Set up your Python virtual environment: ```bash python -m venv labenv ./labenv/Scripts/Activate.ps1 pip install -r requirements.txt azure-ai-projects

   2. Open the .env config file:
    ```bash
    code .env
     ```
   3. Replace your_project_endpoint with your actual Foundry project endpoint.
   
   4. Save (CTRL+S) and close (CTRL+Q) the editor.
   
   ## 🧠 Write Code for the Agent App
   Assicurati di rispettare l’indentazione corretta durante l’inserimento del codice.
   
   1. Apri il file
    ```bash
    code agent.py
     ```
   
   2. Import needed classes
   ```bash
   from azure.identity import DefaultAzureCredential
   from azure.ai.agents import AgentsClient
   from azure.ai.agents.models import FilePurpose, CodeInterpreterTool, ListSortOrder, MessageRole
     ```
   
   3. Connect to the project
   ```bash
   agent_client = AgentsClient(
       endpoint=project_endpoint,
       credential=DefaultAzureCredential(
           exclude_environment_credential=True,
           exclude_managed_identity_credential=True
       )
   )
   with agent_client:
     ```
   
   4. Upload data and create a Code Interpreter Tool
   ```bash
   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])
     ```
   
   5. Define the Agent
   ```bash
   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}")
     ```
   
   6. Create a thread for the conversation
   ```bash
   thread = agent_client.threads.create()
     ```
   
   7. Send the prompt to the Agent
   ```bash
   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)
     ```
   
   8. Check errors
   ```bash
   if run.status == "failed":
       print(f"Run failed: {run.last_error}")
     ```
   
   9. Retrieve the last agent answer
   ```bash
   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}")
     ```
   
   10. Show history of the conversation
   ```bash
   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")
     ```
   
   
   11. Download generated files
   ```bash
   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}")
     ```
   
   12. Delete resources
   ```bash
   agent_client.delete_agent(agent.id)
     ```
   
   ## 🔐 Sign into Azure and Run the App
   1. Autenticate yourself:
   ```bash
   az login 
   

2. Esecute the app:

   python agent.py 
   

3. Insert the prompt:

   "What's the category with the highest cost?"
   "Create a pie chart showing cost by category"
   

4. Continue the conversation, so insert 'quit' to exit.

5. Download the generated files .png:

   download ./<file_name>.png