Note : I used GPT tools to create this code. created this project for learning purpose.

Smart Document Search

Smart Document Search is a Django-based application that provides API endpoints for managing and searching PDF documents using AI-powered text embedding and search capabilities.

Features

API operations support for:

• PDF document upload
• Document search using vector similarity
• Document summarization
• Document deletion

Prerequisites

• Python 3.8+
• Pip (Python package installer)
• OpenAI API key
• Pinecone API key

All other dependencies are listed in requirements.txt and will be installed during the setup process.

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/smart-doc-search.git
   cd smart-doc-search
   

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`
   

3. Install the required packages:

   pip install -r requirements.txt
   

4. Copy the .env.example file to .env and fill in your API keys:

   cp .env.example .env
   

5. Run migrations:

   python manage.py migrate
   

6. Create a documents folder in the project root to store uploaded files.

Usage

1. Start the Django development server:

   python manage.py runserver
   

2. Access the API endpoints:

   • Upload a document: POST /api/upload/
   • Search documents: POST /api/search/
   • Summarize a document: POST /api/summarize/
   • Delete a document: DELETE /api/delete/<document_id>/

API Documentation

Upload a Document

• Endpoint: POST /api/upload/
• Request: Multipart form data with 'file' field containing the PDF document
• Response: JSON object with document details

Search Documents

• Endpoint: POST /api/search/
• Request: JSON object with 'query' field
• Response: List of matching documents with relevance scores

Summarize a Document

• Endpoint: POST /api/summarize/
• Request: JSON object with 'document_id' field
• Response: JSON object with document summary

Delete a Document

• Endpoint: DELETE /api/delete/<document_id>/
• Response: Confirmation of deletion