Extract structured data with Box AI in Python

Box AI exposes intelligent extraction capabilities that enable developers to automatically extract structured key-value pairs from documents through a single API call. This powerful feature transforms unstructured document content into actionable metadata without manual data entry, streamlining document processing workflows for invoices, forms, contracts, and other business documents. This quick start demonstrates how to configure the Box Python SDK, create a metadata template, and use Box AI to extract invoice data and store it as searchable metadata in Box.

Create and configure a Box application

The first step for any Box Platform integration is to create and configure a Box application.

Go to Box Developer Console.
For this quick start, create an App with the Client Credentials Grant application type.
Once the app is created, enable the following scopes:
- Read all files and folders stored in Box
- Write all files and folders stored in Box
- Manage AI

For more information about creating a new Box application, see Create your first application.

Create a Box metadata template

This step requires Admin access to your Box Enterprise. If you do not have access in your current environment, contact your Box administrator.

Box AI enables you to extract data from documents in several ways:

Type	Description	Use case
Freeform extraction	Accepts a string prompt.	Provide a natural language prompt.
Structured extraction with template	Accepts a Box Metadata template key.	Define fields and data types once; simplifies pushing back to Box as metadata.
Structured extraction with fields	Accepts a JSON array of fields.	Run one-off extractions without creating a template.
Enhanced Extract Agent	Uses a specialized agent with a reasoning model.	Use for complex documents or nuanced extraction; works with structured templates or fields.

For this quick start, create a Box metadata template to define the fields you want to extract from your documents. See Customizing Metadata Templates for a detailed walkthrough of the steps to create a Box metadata template in the Box Admin Console.

Give your template a name, for example, Box AI extract quick start.

Create the following fields:

Field name	Type	Description
Client Name	Text	The name of the client receiving the invoice
Invoice Amount	Number	The total amount of the invoice after taxes and fees
Products	Text	The names of the products delivered in the invoice, returned as a comma-delimited string

The field description is used by Box AI to supplement the prompt to the LLM to ensure the right data is extracted.

Click Save to create your template. Make note of the template key to use in a later step.

When you save the template, a list of templates appears. To find the template key, open the template you just created and inspect the URL. The last part of the path is your template key. For example, the URL might look like this: https://app.box.com/master/metadata/templates/boxAiExtractquick start.In this case, the template key is boxAiExtractquick start.

Upload a test file

After preparing the template, select a file to test. For this quick start, use this sample invoice document.

Download the test document, and then drag and drop it into your Box account.
Get the file ID by opening the file in Box and inspecting the URL. The last part of the path is your file ID. For example, the URL might look like this: https://app.box.com/file/2064123286902 In this case, the file ID is 2064123286902.

Configure the environment

Now set up your development environment to run the extraction. For this quick start, use Python and the Box Python SDK version 10. Make sure you have Python 3.11 or higher installed on your machine.

Create a new directory for your project and navigate into it.
Create a virtual environment:
```
python3 -m venv .venv
source .venv/bin/activate
```
Install the Box Python SDK:
```
pip install "boxsdk~=10"
```
Install the python-dotenv package to load environment variables from the .env file:
```
pip install python-dotenv
```
Create an .env file in the root of your project directory and add the following environment variables, replacing the placeholder values with your actual Box app credentials and the IDs from the previous steps:
```
 BOX_DEVELOPER_TOKEN=your_box_developer_token

 BOX_METADATA_TEMPLATE_KEY=your_metadata_template_key
 BOX_FILE_ID=your_box_file_id
```

To get your Developer Token, go to the Box Developer Console, open your app, and navigate to the Configuration tab.

Click Generate Developer Token to create a new token.

For simplicity, this quick start uses a short-lived developer token. In production, you should authenticate using your app’s configured method (for example, Client Credentials Grant) instead of a developer token.

Create the extract.py file

Your development environment is now ready to create the Python script to extract data from the document using Box AI.

Create a new file named extract.py in the root of your project directory and add the following code:

import os

from dotenv import load_dotenv

from box_sdk_gen import (
    AiItemBase,
    BoxClient,
    BoxDeveloperTokenAuth,
    CreateAiExtractStructuredMetadataTemplate,
    CreateAiExtractStructuredMetadataTemplateTypeField,
    CreateFileMetadataByIdScope
)

load_dotenv()

developer_token = os.getenv("BOX_DEVELOPER_TOKEN")
file_id = os.getenv("BOX_FILE_ID")
template_key = os.getenv("BOX_METADATA_TEMPLATE_KEY")

def get_box_client(token: str) -> BoxClient:
    
    if not developer_token:
        raise ValueError("BOX_DEVELOPER_TOKEN is not set in environment variables.")
    
    auth = BoxDeveloperTokenAuth(token=token)
    client = BoxClient(auth=auth)

    return client

def main():
    client = get_box_client(token=developer_token)

    me = client.users.get_user_me()
    print(f"My user ID is {me.id}")


if __name__ == "__main__":
    main()

This code loads the environment variables from the .env file, initializes the Box SDK client, and prints the current user’s ID to validate that the client is working correctly.

Run the script using the following command in your terminal:
```
python extract.py
```
If the Box SDK client is set up correctly, the console displays your user ID. For example:
```
My user ID is 123456789
```

Extract data

With a working Box SDK client, you can add the code to extract data from the document using Box AI.

Between the get_box_client function and the main function, add the following function:

def extract_metadata(client: BoxClient, file_id: str, template_key: str) -> dict:
    metadata = client.ai.create_ai_extract_structured(
        [AiItemBase(id=file_id)],
        metadata_template=CreateAiExtractStructuredMetadataTemplate(
            template_key=template_key,
            type=CreateAiExtractStructuredMetadataTemplateTypeField.METADATA_TEMPLATE,
            scope="enterprise",
        ),
    )
    
    return metadata.to_dict()['answer']

This function uses the Box AI create_ai_extract_structured method to extract metadata from the specified file. Your BoxClient, the file ID, and the metadata template key created earlier are sent to the function, which returns the extracted metadata as a dictionary.

Add the function call to extract the metadata in the main function. Ensure that the new main function contains the following logic:
```
def main():
    client = get_box_client(token=developer_token)

    me = client.users.get_user_me()
    print(f"My user ID is {me.id}")

    metadata = extract_metadata(client=client, file_id=file_id, template_key=template_key)

    print(f"Extracted Metadata: {metadata}")
```
The SDK handles the API call to Box AI and returns the extracted metadata as an AiExtractStructuredResponse object. In this quick start, the code converts this object to a dictionary and returns the answer field that contains the extracted key/value pairs.

Print out the extracted metadata to the console to verify that the extraction was successful by running the following command in your terminal:

python extract.py

If the extraction was successful, the console displays your user ID followed by the extracted metadata from the invoice document.

My user ID is 123456789

Extracted Metadata: {'clientName': 'ACME Inc', 'invoiceAmount': 1106.06, 'products': 'Polyol, Diisocyanate, 
Carbon Dioxide, Laser, Lens, Oleic Acid, Glycerine, Sodium Tallowate, Paint Base, Polypropylene, Rubber, 
Additive, Pigment, Aluminum Silicate, Magnesium Silicate, Zinc Oxide, Distilled Solvent, Petroleum Distillate, 
Sulfur Dioxide, Sodium Benzoate, Dust cap, Ferrite cap, Cone and coil assembly, Cleaner, Polypropylene pellets, 
Polypropylene chips, Polypropylene blocks, Polypropylene slag, Parts Wash Solvent, Jar, 
Plastic Bottle - 15.2 FL Oz (450 ml), Polymer'}

Add Box metadata to the file

Now that you have extracted metadata from the document, you can use these key/value pairs in your application: push to databases, integrate with CRMs, feed to agents for processing, or trigger automated workflows.This quick start demonstrates pushing the extracted data back to Box as file metadata. Box metadata management enables powerful filtering and search capabilities across your content. For example, you can query all invoices over $500 from the last 30 days, create dashboards in Box Apps, or surface key document insights directly in the Box web application.

Push the extracted metadata back to Box by adding the following function between the extract_metadata function and the main function:
```
def push_metadata(client: BoxClient, file_id: str, metadata: dict, template_key: str) -> dict:
    attached_metadata = client.file_metadata.create_file_metadata_by_id(
        file_id,
        CreateFileMetadataByIdScope.ENTERPRISE,
        template_key,
        metadata,
    )
    return attached_metadata.to_dict()
```
This function uses the create_file_metadata_by_id method to attach metadata to the specified file, processing the BoxClient, file ID, metadata dictionary, and template key. The API itself returns a MetadataFull object. The function converts this object to a dictionary and returns it.

Add the function call to push the metadata in the main function. Ensure that the updated main function contains the following logic:

def main():
    client = get_box_client(token=developer_token)

    me = client.users.get_user_me()
    print(f"My user ID is {me.id}")

    metadata = extract_metadata(client=client, file_id=file_id, template_key=template_key)

    print(f"Extracted Metadata: {metadata}")

    attached_metadata = push_metadata(client=client, file_id=file_id, metadata=metadata, template_key=template_key)

    print(f"Attached Metadata: {attached_metadata}")

Run the following command in your terminal:

python extract.py

If the script is successful, the console displays your user ID, the extracted metadata, and the attached metadata response from Box. For example:

My user ID is 123456789

Extracted Metadata: {'clientName': 'ACME Inc', 'invoiceAmount': 1106.06, 'products': 'Polyol, Diisocyanate,
Carbon Dioxide, Laser, Lens, Oleic Acid, Glycerine, Sodium Tallowate, Paint Base, Polypropylene, Rubber,
Additive, Pigment, Aluminum Silicate, Magnesium Silicate, Zinc Oxide, Distilled Solvent, Petroleum Distillate,
Sulfur Dioxide, Sodium Benzoate, Dust cap, Ferrite cap, Cone and coil assembly, Cleaner, Polypropylene pellets,
Polypropylene chips, Polypropylene blocks, Polypropylene slag, Parts Wash Solvent, Jar,
Plastic Bottle - 15.2 FL Oz (450 ml), Polymer'}

Attached Metadata: {'invoiceAmount': 1106.06, 'products': 'Polyol, Diisocyanate, Carbon Dioxide, Laser, Lens, 
Oleic Acid, Glycerine, Sodium Tallowate, Paint Base, Polypropylene, Rubber, Additive, Pigment, Aluminum Silicate, 
Magnesium Silicate, Zinc Oxide, Distilled Solvent, Petroleum Distillate, Sulfur Dioxide, Sodium Benzoate, Dust cap, 
Ferrite cap, Cone and coil assembly, Cleaner, Polypropylene pellets, Polypropylene chips, Polypropylene blocks, 
Polypropylene slag, Parts Wash Solvent, Jar, Plastic Bottle - 15.2 FL Oz (450 ml), Polymer', 
'clientName': 'ACME Inc', '$parent': 'file_1956534287859', '$template': 'boxAiExtractquick start', 
'$scope': 'enterprise_899905961', '$version': 0, '$canEdit': True, '$id': '4d4f0b55-d45a-4ba4-9ff6-1241182cb76a', 
'$type': 'boxAiExtractquick start-c4024235-2384-49f4-9286-ada6d68fd6a9', '$typeVersion': 0, 'extra_data': 
{'invoiceAmount': 1106.06, 'products': 'Polyol, Diisocyanate, Carbon Dioxide, Laser, Lens, Oleic Acid, Glycerine, 
Sodium Tallowate, Paint Base, Polypropylene, Rubber, Additive, Pigment, Aluminum Silicate, Magnesium Silicate, 
Zinc Oxide, Distilled Solvent, Petroleum Distillate, Sulfur Dioxide, Sodium Benzoate, Dust cap, Ferrite cap, 
Cone and coil assembly, Cleaner, Polypropylene pellets, Polypropylene chips, Polypropylene blocks, Polypropylene slag, 
Parts Wash Solvent, Jar, Plastic Bottle - 15.2 FL Oz (450 ml), Polymer', 'clientName': 'ACME Inc'}}

​Resources

Resources