Chat with your PDF: Seamlessly interact with documents using Amazon Bedrock, RAG, S3, Langchain, and Streamlit for intelligent, responsive conversations.

Architecture

PDFPal is a web application designed for processing PDF files. It offers two main components:

• Admin Interface: For processing PDFs, creating a vector store, and uploading it to Amazon S3.
• User Interface: For interacting with the processed PDFs and vector store.

Getting Started

Follow the instructions below to set up and run both the Admin and User components of the application.

Prerequisites

• Python 3.11 or later (if not using Docker)
• Docker (if using Docker)
• Amazon S3 account and bucket for storing vector store files
• AWS CLI installed for configuring AWS credentials

Configuration

1. Configure AWS CLI:

   Ensure you have AWS CLI installed and configured. Run the following command to set up your AWS credentials:

   aws configure

You will need to provide your AWS Access Key ID, Secret Access Key, region, and output format.

Setup Instructions

For Both User and Admin

1. Clone the Repository:

git clone https://github.com/your-repository-url
cd your-repository-directory

2. Install Dependencies:

Create a requirements.txt file with the following content:

Copy code
boto3
streamlit
faiss-cpu
langchain
langchain-community

Then, install the dependencies using:

pip install -r requirements.txt

Using Docker

1. Build Docker Images:

For the Admin Interface:

docker build -t pdfpal-admin .

For the User Interface:

docker build -t pdfpal-user .

2. Run Docker Containers:

Admin Interface

For the Admin Interface (accessible at http://localhost:8083):

docker run -p 8083:8083 pdfpal-admin

Client Interface

For the User Interface (accessible at http://localhost:8084):

docker run -p 8084:8084 pdfpal-user

Using the Application

Admin Interface:

• Upload a PDF: Use the file uploader widget to select and upload a PDF file.
• Process PDF: The application will process the PDF, split its content, create a vector store, and upload it to S3.

User Interface:

• Interact with Processed PDFs: Access the processed PDFs and vector store as needed.

Troubleshooting:

• No Response from App: Ensure the application is running by checking the terminal output.
• File Upload Issues: Make sure the file is in PDF format.
• S3 Errors: Verify AWS credentials and S3 bucket permissions.