Streamlining Document Processing with Azure - Doc Intelligence in-a-Box

In today's data-driven world, organizations often grapple with vast amounts of unstructured data embedded within documents. Manual extraction of this information is not only time-consuming but also prone to errors. To address this challenge, Microsoft has introduced the "Doc Intelligence in-a-Box" solution, leveraging Azure AI Document Intelligence to automate and streamline document data extraction.

Key Features

• Automated Data Extraction: Utilizes Azure AI Document Intelligence to accurately extract text, key-value pairs, tables, and other relevant data from PDF forms.
• Seamless Data Storage: Integrates with Azure Cosmos DB to store extracted data, facilitating easy access and management.
• Rapid Deployment: As part of the AI-in-a-Box framework, the solution is designed for quick implementation, reducing the time required to operationalize AI and ML projects.

How It Works

The solution employs Azure AI Document Intelligence, a cloud-based service that combines Optical Character Recognition (OCR) and advanced machine learning models. This enables the extraction of structured and unstructured data from various document formats, including PDFs. Once the data is extracted, it is stored in Azure Cosmos DB, a globally distributed, multi-model database service that offers scalability and flexibility.

Benefits

• Enhanced Efficiency: Automating the data extraction process reduces manual effort, allowing teams to focus on more strategic tasks.
• Improved Accuracy: Advanced AI models minimize errors associated with manual data entry, ensuring high data quality.
• Scalability: The solution is designed to handle large volumes of documents, making it suitable for organizations of all sizes.
• Flexibility: Supports various document formats and integrates seamlessly with existing workflows and systems.

The data extraction process in the "Doc Intelligence in-a-Box" solution involves several key steps:

Upload PDFs:

• You start by uploading PDF documents to a designated container in Azure Data Lake Storage Gen2 (ADLS Gen2).

Trigger Processing:

• An Azure Logic App is triggered when a PDF is uploaded. This Logic App sends the PDF file location to an Azure Functions app for processing.

Split PDF into Pages:

• The Azure Functions app splits the PDF into single pages if it contains multiple pages. Each page is saved as an individual PDF file in ADLS Gen2.

Send to Azure AI Document Intelligence:

• The single-page PDF files are sent to Azure AI Document Intelligence via a REST API (HTTPS POST) for data extraction.

Extract Data:

• Azure AI Document Intelligence processes the PDF files and extracts the relevant data. This includes text, tables, and other structured information.

Store Extracted Data:

• The extracted data is then stored in Azure Cosmos DB for further use and analysis.

Getting Started

To explore and implement the "Doc Intelligence in-a-Box" solution, visit the official GitHub repository: GitHub

The repository provides comprehensive documentation, including setup instructions, prerequisites, and sample code to assist in deployment.

By adopting the "Doc Intelligence in-a-Box" solution, organizations can transform their document processing workflows, leading to increased productivity and more informed decision-making.

Clone the Repository:

git clone https://github.com/Azure-Samples/doc-intelligence-in-a-box.git
cd doc-intelligence-in-a-box        

Set Up Azure Resources:

• Ensure you have an Azure subscription.
• Create the necessary Azure resources such as Azure Key Vault, Azure Blob Storage, Azure Cosmos DB, and Azure Functions. For detailed instructions, you can refer to the README.md file in the repository

Run the Solution

To test the "Doc Intelligence in-a-Box" solution after deployment, follow these steps:

1. Upload PDFs to Azure Data Storage:
2. Trigger the Processing from the logic app
3. Verify the Results
4. Review the result in Cosmos DB

References

https://github.com/Azure-Samples/doc-intelligence-in-a-box