PDF-QueryBOT

PDF-QueryBOT is a full-stack application designed to allow users to upload PDF documents and interact with them through natural language processing. Users can ask questions regarding the content of the uploaded PDFs, and the backend processes these documents to provide answers.

Application Architecture

The application is built using a microservices architecture, with the following components:

• Frontend: Developed with React.js, providing a user interface for document upload and interaction.
• Backend: Built with FastAPI, handling PDF uploads, question processing, and answer retrieval.
• NLP Processing: Utilizes LangChain for processing natural language questions and generating answers from PDF content.

Setup Instructions

1. Clone the repository:

   git clone https://github.com/suhailskhan9/PDF-QueryBOT.git
   cd PDF-QueryBOT
   

2. Install dependencies:
   • For the backend (inside the backend directory):

   pip install -r requirements.txt

   • For the frontend (inside the frontend directory):

   npm install
   

3. Start the backend server:

   uvicorn main:app --reload
   

4. Start the frontend application:

   npm start
   

API Documentation

1. Upload PDFs

Uploads PDF files, extracts text, splits it into chunks, and stores vector embeddings for retrieval.

Endpoint

POST /upload

Request

• Content-Type: multipart/form-data
• Body Parameters:

Parameter	Type	Required	Description
files	List[UploadFile]	✅ Yes	List of PDF files to upload

Response

• Success (200 OK)

  {
    "message": "PDFs uploaded and processed successfully"
  }

2. Ask a Question

Queries the stored PDFs for an answer based on stored embeddings.

Endpoint

POST /ask

Request

• Content-Type: application/json
• Body Parameters:

Parameter	Type	Required	Description
question	str	✅ Yes	The question to ask based on uploaded PDFs

Response

• Success (200 OK)

  {
    "answer": "Summarized response from the PDFs"
  }