Adding Knowledge with File Search

In the previous chapters, you created a basic agent and gave it instructions through a system prompt.
Now it's time to make your agent smarter by grounding it in your own data.

Why Add Knowledge?

By default, the model only knows what it was trained on - it doesn't have access to your organization's private or domain-specific information.
To bridge this gap, we'll use Retrieval-Augmented Generation (RAG).

• RAG lets the agent fetch relevant information from your own data before generating a response.
• This ensures your agent's answers are accurate, up-to-date, and grounded in real information.
• In Microsoft Foundry, we'll use the File Search feature to implement this.

In this chapter, you'll use a folder called ./documents that contains information about Contoso Pizza stores - such as locations, opening hours, and menus.

We'll upload these files to a vector store and connect that store to the agent using a File Search tool.

Step 1 - Update Your Imports

First, update your imports at the top of agent.py to include the necessary classes:

import os
import glob
from dotenv import load_dotenv
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient
from azure.ai.projects.models import PromptAgentDefinition, FileSearchTool, Tool

Step 2 - Create the Vector Store

Add this code after creating your openai_client to create a vector store and upload documents:

vector_store_id = ""  # Set to your vector store ID if you already have one

## -- FILE SEARCH -- ##

if vector_store_id:
    vector_store = openai_client.vector_stores.retrieve(vector_store_id)
    print(f"Using existing vector store (id: {vector_store.id})")
else:
    # Create vector store for file search
    vector_store = openai_client.vector_stores.create(name="ContosoPizzaStores")
    print(f"Vector store created (id: {vector_store.id})")

    # Upload file to vector store
    for file_path in glob.glob("documents/*.md"):
        file = openai_client.vector_stores.files.upload_and_poll(
            vector_store_id=vector_store.id, file=open(file_path, "rb")
        )
        print(f"File uploaded to vector store (id: {file.id})")
## -- FILE SEARCH -- ##

Why:

• A vector store stores and indexes document embeddings for semantic search.
• The first time you run the script, it creates a new vector store and uploads all documents.
• On subsequent runs, you can set vector_store_id to reuse the existing store.

Tip: After the first run, copy the vector store ID from the output and set it in the vector_store_id variable to avoid recreating it each time.

Step 3 - Add the File Search Tool

Now add the File Search tool to your agent's toolset. Add this before creating the agent:

## Define the toolset for the agent
toolset: list[Tool] = []
toolset.append(FileSearchTool(vector_store_ids=[vector_store.id]))

Step 4 - Update the Agent Creation

Modify your agent creation to include the toolset:

## Create a Foundry Agent
agent = project_client.agents.create_version(
    agent_name="hello-world-agent",
    definition=PromptAgentDefinition(
        model=os.environ["MODEL_DEPLOYMENT_NAME"],
        instructions=open("instructions.txt").read(),
        tools=toolset,
    ),
)
print(f"Agent created (id: {agent.id}, name: {agent.name}, version: {agent.version})")

Step 5 - Run the Agent

Try it out:

python agent.py

Ask questions like:

"Which Contoso Pizza stores are open after 8pm?"
"Where is the nearest Contoso Pizza store?"

Type exit or quit to stop the conversation.

Recap

In this chapter, you:

• Learned how RAG grounds your agent with your own data
• Created and populated a vector store directly in your script
• Added a File Search tool to your agent
• Extended your PizzaBot to answer questions about Contoso Pizza stores

Final code sample

import os
import glob
from dotenv import load_dotenv
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient
from azure.ai.projects.models import PromptAgentDefinition, FileSearchTool, Tool

load_dotenv()

vector_store_id = ""  # Set to your vector store ID if you already have one

## Configure Project Client
project_client = AIProjectClient(
    endpoint=os.environ["PROJECT_ENDPOINT"],
    credential=DefaultAzureCredential(),
)
openai_client = project_client.get_openai_client()


## -- FILE SEARCH -- ##

if vector_store_id:
    vector_store = openai_client.vector_stores.retrieve(vector_store_id)
    print(f"Using existing vector store (id: {vector_store.id})")
else:
    # Create vector store for file search
    vector_store = openai_client.vector_stores.create(name="ContosoPizzaStores")
    print(f"Vector store created (id: {vector_store.id})")

    # Upload file to vector store
    for file_path in glob.glob("documents/*.md"):
        file = openai_client.vector_stores.files.upload_and_poll(
            vector_store_id=vector_store.id, file=open(file_path, "rb")
        )
        print(f"File uploaded to vector store (id: {file.id})")
## -- FILE SEARCH -- ##


## Define the toolset for the agent
toolset: list[Tool] = []
toolset.append(FileSearchTool(vector_store_ids=[vector_store.id]))


## Create a Foundry Agent
agent = project_client.agents.create_version(
    agent_name="hello-world-agent",
    definition=PromptAgentDefinition(
        model=os.environ["MODEL_DEPLOYMENT_NAME"],
        instructions=open("instructions.txt").read(),
        tools=toolset,
    ),
)
print(f"Agent created (id: {agent.id}, name: {agent.name}, version: {agent.version})")


## Create a conversation for the agent interaction
conversation = openai_client.conversations.create()
print(f"Created conversation (id: {conversation.id})")

## Chat with the agent

while True:
    # Get the user input
    user_input = input("You: ")

    if user_input.lower() in ["exit", "quit"]:
        print("Exiting the chat.")
        break

    # Get the agent response
    response = openai_client.responses.create(
        conversation=conversation.id,
        input=user_input,
        extra_body={"agent": {"name": agent.name, "type": "agent_reference"}},
    )

    # Print the agent response
    print(f"Assistant: {response.output_text}")