FileFlow AI is an intelligent document classification system designed to streamline the document handling workflow for institutions using a combination of OCR (Optical Character Recognition) technology and Cosmos DB to store and retrieve documents. The solution is built with Azure AI integration for OCR and storage functionality leveraging Azure Blob Storage and Cosmos DB for managing user account data and their associated documents.
Core Capabilities:
- OCR Text Extraction: Utilizing Azure’s prebuilt document models, we extract text from images and PDFs for further processing.
- Document Storage: Secure storage of documents in Azure Blob Storage with metadata associated in Cosmos DB.
- User Account Association: Accounts are linked with documents, ensuring easy access and retrieval based on user data like names.
- 🖥️ Document Classification: Extracts document type and extracts user-specific data (e.g., names).
- 📑 OCR Integration: Uses Azure AI Document Intelligence to extract structured text from documents.
- 🔐 Cosmos DB Storage: All account data and document links are securely stored in Azure Cosmos DB.
- 🔗 Seamless Document Upload: Upload documents to Azure Blob Storage and link them to user accounts in Cosmos DB.
- 📡 Account Lookup and Linking: Search by account number and manage linked documents.
Ensure the following are installed:
- Python 3.8+
- pip (Python package installer)
- Streamlit for the frontend.
-
Clone the Repository:
git clone https://github.com/mohammeduvez29/FileFlowAI/cd FileFlowAI -
Set Up Virtual Environment:
python -m venv venvsource venv/bin/activate # On Windows: venv\Scripts\activate -
Install Dependencies:
pip install -r requirements.txt -
Run the Application:
python app.pyThis will start the Streamlit application and allow you to upload documents, extract text, and link them to accounts.
-
Launch the application using:
streamlit run app.py -
In the Streamlit UI, you will be presented with two options:
- Upload Document: Upload a document to analyze its contents using Azure’s OCR and either create a new account or link it to an existing one.
- Search by Account Number: Search and view documents already linked to an existing account.
-
The app will display:
- Extracted document type.
- Extracted name (if available).
- Options to either link to an existing account or create a new one.
-
Account Management:
- Users can link documents to existing accounts based on extracted names.
- If no account is found, the system allows you to create a new account and associate the document with it.
- Streamlit: For building the web interface where users can upload documents and view results.
- Azure Form Recognizer (OCR): For extracting text from documents.
- Azure Blob Storage: For storing the uploaded documents securely.
- Azure Cosmos DB: For storing user account information and linked document metadata.
- Azure Key Vault: (Optional) For securely managing credentials and keys.
FileFlowAI/
├── app.py # Main Streamlit application
├── config.py # Environment configuration and Azure credentials setup
├── azure_blob.py # Logic for interacting with Azure Blob Storage
├── azure_ocr.py # OCR text extraction logic using Azure AI
├── requirements.txt # Required Python dependencies
├── .env # Environment variables for configuration (BLOB_CONNECTION_STRING, COSMOS_URL, etc.)
- Python 3.8+
- Streamlit
- azure-ai-formrecognizer
- azure-storage-blob
- azure-cosmos
- python-dotenv for managing environment variables
You can install all required dependencies by running:
pip install -r requirements.txt
- Users can upload a document (PDF, JPG, PNG).
- The system uses Azure OCR to extract text and recognize the document type (e.g., PAN Card).
- The extracted data (such as name) is used to check if an account already exists in Cosmos DB.
- If the account exists, the document is linked to the existing account.
- If no account exists, the user can create a new account, and the document will be linked to it.