Skip to main content

Quick Start - 5 Minute Tutorial

Get started with DocuDevs in just 5 minutes! This tutorial will walk you through processing your first document and extracting structured data using AI.

What You'll Learn​

By the end of this tutorial, you'll know how to:

  • Install and configure the DocuDevs SDK
  • Upload and process a document
  • Extract structured data using AI
  • Retrieve and use the results

Prerequisites​

  • Python 3.9+ installed on your system
  • A DocuDevs API key (contact us to get started)
  • A sample document (PDF, Word, or image file)

Step 1: Install the SDK (30 seconds)​

pip install docu-devs-api-client

Step 2: Set Your API Key (10 seconds)​

export API_KEY="your-api-key-here"

Step 3: Process Your First Document (2 minutes)​

Create a simple Python script to process a document:

import os
import asyncio
from docudevs.docudevs_client import DocuDevsClient

async def process_document():
    # Initialize client
    client = DocuDevsClient(token=os.getenv('API_KEY'))
    
    # Read your document
    with open("your-document.pdf", "rb") as f:
        document_data = f.read()
    
  # Submit for AI extraction (non-OCR)
  job_id = await client.submit_and_process_document(
    document=document_data,
    document_mime_type="application/pdf",
    prompt="Extract all key information from this document"
  )
    print(f"Document submitted for processing. Job ID: {job_id}")
    
    # Wait for processing to complete; request canonical JSON output
    data = await client.wait_until_ready(job_id, result_format="json")
    print("Processing complete!")
    
  # Display results as JSON (extraction returns structured data)
  import json
  print("\nExtracted Data:")
  print(json.dumps(data, indent=2))

# Run the processing
asyncio.run(process_document())

Step 4: Run and See Results (2 minutes)​

Save the script as quick_start.py and run it:

python quick_start.py

You'll see output like:

Document submitted for processing. Job ID: abc123-def456-ghi789
Processing complete!

Extracted Data:
{
  "invoice_number": "INV-2024-001",
  "date": "2024-01-15",
  "total_amount": "$1,250.00",
  "vendor": "Acme Corp",
  "items": [
    {
      "description": "Professional Services",
      "quantity": 10,
      "unit_price": "$125.00"
    }
  ]
}

What Just Happened?​

  1. Document Upload: Your document was securely uploaded to DocuDevs
  2. AI Processing: Our AI analyzed the document and extracted structured data
  3. Smart Extraction: The AI understood the document type and extracted relevant fields
  4. JSON Output: Results were returned in clean, structured JSON format

Bonus: AI-Powered Document Analysis (1 minute)​

Want to go beyond structured extraction? Try using AI to understand and summarize your documents:

async def analyze_document_with_ai():
    client = DocuDevsClient(token=os.getenv('API_KEY'))
    
    # Read your document
    with open("your-document.pdf", "rb") as f:
        document_data = f.read()
    
    # First, process with OCR to extract text
    ocr_job_id = await client.submit_and_ocr_document(
        document=document_data,
        document_mime_type="application/pdf",
        ocr="PREMIUM",
        ocr_format="markdown"
    )
    
    # Wait for OCR to complete
    await client.wait_until_ready(ocr_job_id)
    
    # Now ask AI to analyze the document
    analysis = await client.submit_and_wait_for_generative_task(
        parent_job_id=ocr_job_id,
        prompt="Summarize this document and explain what type of document it is and its main purpose."
    )
    
  # Display the AI response
  print("\nAI Analysis:")
  print(analysis.generated_text)

# Run the analysis
asyncio.run(analyze_document_with_ai())

This approach is perfect for:

  • Document Summaries: Get quick overviews of long documents
  • Document Classification: Automatically categorize document types
  • Question Answering: Ask specific questions about document content
  • Content Transformation: Convert documents to different formats or styles

Next Steps - Choose Your Path​

🚀 For Developers​

📋 For Business Users​

  • Use Cases - See real-world application examples
  • Cases - Organize documents into collections
  • Templates - Create reusable document templates

⚙️ For System Integrators​

Try Different Document Types​

DocuDevs works with various document formats:

instruction = "Extract invoice number, date, vendor, total amount, and all line items"

Common Customizations​

Add Custom Schema​

schema = {
    "invoice_number": "The invoice number",
    "date": "Invoice date in YYYY-MM-DD format",
    "vendor": "Company that issued the invoice",
    "total": "Total amount as number",
    "currency": "Currency code (USD, EUR, etc.)"
}

job_id = await client.submit_and_process_document(
  document=document_data,
  document_mime_type="application/pdf",
  prompt="Extract invoice data according to the schema",
  schema=schema
)

Process Multiple Documents​

# Create a case for organizing related documents
case_response = await client.create_case({
    "name": "Q4 2024 Invoices",
    "description": "All vendor invoices for Q4 processing"
})
case_id = case_response.parsed.id

# Upload multiple documents to the case
for filename in ["invoice1.pdf", "invoice2.pdf", "invoice3.pdf"]:
    with open(filename, "rb") as f:
        await client.upload_case_document(
            case_id=case_id,
            document=f.read(),
            filename=filename
        )

Getting Help​

What's Next?​

You've successfully processed your first document! Now you're ready to:

  1. Explore Advanced Features: Learn about Cases for document organization
  2. Customize Processing: Dive into Templates and Configuration
  3. Build Integration: Review Best Practices for production use
  4. Scale Your Solution: Discover performance optimization techniques

Ready to build something amazing with DocuDevs? Let's continue with the detailed first document guide!