Documind is an advanced document processing tool that leverages AI to extract structured data from PDFs. It is built to handle PDF conversions, extract relevant information, and format results as specified by customizable schemas.

• Converts PDFs to images for detailed AI processing.
• Uses OpenAI’s API to extract and structure information.
• Allows users to specify extraction schemas for various document formats.
• Designed for flexible deployment on local or cloud environments.

A demo of the documind hosted version will be available soon for you to try out! The hosted version provides a seamless experience with fully managed APIs, so you can skip the setup and start extracting data right away.

For full access to the hosted service, please request access and we’ll get you set up.

Before using documind, ensure the following software dependencies are installed:

• Ghostscript: documind relies on Ghostscript for handling certain PDF operations.
• GraphicsMagick: Required for image processing within document conversions.

Install both on your system before proceeding:

# On macOS
brew install ghostscript graphicsmagick

# On Debian/Ubuntu
sudo apt-get update
sudo apt-get install -y ghostscript graphicsmagick

Ensure Node.js (v18+) and NPM are installed on your system.

You can install documind via npm:

npm install documind

documind requires an .env file to store sensitive information like API keys and Supabase configurations.

Create an .env file in your project directory and add the following:

OPENAI_API_KEY=your_openai_api_key
SUPABASE_URL=your_supabase_url
SUPABASE_KEY=your_supabase_key
SUPABASE_BUCKET=your_supabase_bucket_name

First, import documind and define your schema. The schema outline what information documind should look for in each document. Here’s a quick setup to get started.

The schema is an array of objects where each object defines:

• name: Field name to extract.
• type: Data type (e.g., "string", "number", "array", "object").
• description: Description of the field.
• children (optional): For arrays and objects, define nested fields.

Example schema for a bank statement:

const schema = [
  {
    name: "accountNumber",
    type: "string",
    description: "The account number of the bank statement."
  },
  {
    name: "openingBalance",
    type: "number",
    description: "The opening balance of the account."
  },
  {
    name: "transactions",
    type: "array",
    description: "List of transactions in the account.",
    children: [
      {
        name: "date",
        type: "string",
        description: "Transaction date."
      },
      {
        name: "creditAmount",
        type: "number",
        description: "Credit Amount of the transaction."
      },
      {
        name: "debitAmount",
        type: "number",
        description: "Debit Amount of the transaction."
      },
      {
        name: "description",
        type: "string",
        description: "Transaction description."
      }
    ]
  },
  {
    name: "closingBalance",
    type: "number",
    description: "The closing balance of the account."
  }
];

Use documind to process a PDF by passing the file URL and the schema.

import { extract } from 'documind';

const runExtraction = async () => {
  const result = await extract({
    file: 'https://bank_statement.pdf',
    schema
  });

  console.log("Extracted Data:", result);
};

runExtraction();

Here’s an example of what the extracted result might look like:

 {
  "success": true,
  "pages": 1,
  "data": {
    "accountNumber": "100002345",
    "openingBalance": 3200,
    "transactions": [
        {
        "date": "2021-05-12",
        "creditAmount": null,
        "debitAmount": 100,
        "description": "transfer to Tom" 
      },
      {
        "date": "2021-05-12",
        "creditAmount": 50,
        "debitAmount": null,
        "description": "For lunch the other day"
      },
      {
        "date": "2021-05-13",
        "creditAmount": 20,
        "debitAmount": null,
        "description": "Refund for voucher"
      },
      {
        "date": "2021-05-13",
        "creditAmount": null,
        "debitAmount": 750,
        "description": "May's rent"
      }
    ],
    "closingBalance": 2420
  },
  "fileName": "bank_statement.pdf",
}

Contributions are welcome! Please submit a pull request with any improvements or features.

This project is licensed under the AGPL v3.0 License.