Skip to main content

Named Configurations

Save and reuse processing settings, prompts, and schemas across multiple document processing requests.

Overview​

Named configurations allow you to define a set of processing parameters once and reuse them by name. This ensures consistency across your application and simplifies your API calls.

Key Benefits:

  • Consistency: Ensure all documents of a certain type are processed with the exact same settings.
  • Schema Management: Centralize your JSON schemas to guarantee consistent output formats.
  • Simplicity: Pass a single configuration name instead of a complex JSON object with every request.

Managing Configurations​

Saving a Configuration​

To create or update a configuration, you define the processing parameters (OCR, LLM, Prompt, Schema) and save them under a unique name.

from docudevs.docudevs_client import DocuDevsClient
from docudevs.models import UploadCommand
import os
import asyncio

client = DocuDevsClient(token=os.getenv('API_KEY'))

async def save_invoice_config():
    # Define the configuration using UploadCommand
    config = UploadCommand(
        prompt="Extract invoice data including vendor details, line items, and totals.",
        schema={
            "type": "object",
            "properties": {
                "vendor_name": {"type": "string"},
                "invoice_number": {"type": "string"},
                "total_amount": {"type": "number"},
                "date": {"type": "string", "format": "date"}
            },
            "required": ["vendor_name", "total_amount"]
        },
        ocr="PREMIUM",
        llm="HIGH",
        extraction_mode="STEPS",
        barcodes=True
    )

    # Save the configuration with the name "invoice-config"
    result = await client.save_configuration("invoice-config", config)
    print(f"Configuration saved: {result}")

# asyncio.run(save_invoice_config())

Listing Configurations​

Retrieve a list of all saved configurations in your organization.

configs = await client.list_configurations()
for config in configs:
    print(f"Name: {config.name}, Created: {config.created_at}")

Retrieving a Configuration​

Get the details of a specific configuration to inspect its settings.

config = await client.get_configuration("invoice-config")
print(f"Prompt: {config.prompt}")

Deleting a Configuration​

Remove a configuration when it is no longer needed.

await client.delete_configuration("invoice-config")

Using Configurations​

Once saved, you can use a configuration to process documents. This applies all the settings defined in the configuration to the processing job.

with open("invoice.pdf", "rb") as f:
    document_data = f.read()

# Process document using the saved configuration
job_guid = await client.submit_and_process_document_with_configuration(
    document=document_data,
    document_mime_type="application/pdf",
    configuration_name="invoice-config"
)

# Wait for result
result = await client.wait_until_ready(job_guid)
print(result)

Configuration Parameters​

When creating a configuration, you can specify the following parameters:

ParameterTypeDescription
promptStringCustom instructions for data extraction.
schemaObjectJSON Schema defining the exact output structure.
ocrEnumOCR quality level (NONE, DEFAULT, PREMIUM).
llmEnumLLM model quality (DEFAULT, MINI, HIGH).
extractionModeEnumExtraction method (SIMPLE, STEPS).
barcodesBooleanEnable barcode and QR code scanning.
describeFiguresBooleanEnable AI description of figures and images.

Best Practices​

  • Version Your Configurations: If you change your schema significantly, consider creating a new configuration (e.g., invoice-v2) to avoid breaking existing integrations.
  • Use Schemas: Always define a strict JSON schema to ensure predictable API responses.
  • Test Prompts: Experiment with different prompts in the playground or via temporary configurations before finalizing them.
  • Optimize Costs: Use DEFAULT OCR and LLM for simple documents and HIGH only when necessary.