https://github.com/didibrabosa/ai-pdf-reader

AI-powered PDF Q&A system using FastAPI, ChromaDB, and OpenAI. Extracts, indexes, and retrieves relevant text chunks to answer questions.
https://github.com/didibrabosa/ai-pdf-reader

ai chromadb chunks dotenv fastapi gpt-4 langchain openai pydantic pypdf2 python uvicorn

Last synced: 1 day ago
JSON representation

AI-powered PDF Q&A system using FastAPI, ChromaDB, and OpenAI. Extracts, indexes, and retrieves relevant text chunks to answer questions.

• Host: GitHub
• URL: https://github.com/didibrabosa/ai-pdf-reader
• Owner: didibrabosa
• Created: 2025-01-29T19:46:38.000Z (12 days ago)
• Default Branch: main
• Last Pushed: 2025-02-07T01:15:04.000Z (4 days ago)
• Last Synced: 2025-02-07T02:22:52.408Z (4 days ago)
• Topics: ai, chromadb, chunks, dotenv, fastapi, gpt-4, langchain, openai, pydantic, pypdf2, python, uvicorn
• Language: Python
• Homepage: https://github.com/Piyush150398/RAG_Application
• Size: 1.17 MB
• Stars: 0
• Watchers: 1
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# RAG PDF Reader

## Project Description

The **RAG PDF Reader** is an application that implements a Retrieval-Augmented Generation (**RAG**) system using **FastAPI**. The app reads a specific PDF file ("Harry Potter AI.pdf"), extracts its text, splits it into **chunks**, and stores these chunks in a vector database with **ChromaDB**. When you ask a question, the system retrieves the most relevant text chunks and uses an **OpenAI** model to generate an answer.

### How It Works

1. Startup Process:

   - The PDF file is read and its text is extracted.

   - The extracted text is split into chunks.

   - The chunks are stored in a ChromaDB collection.

   - The response is returned to the user via a **FastAPI**.

2. Query Processing:

   - A POST request is sent to the /ask_ai endpoint with a question.

   - The system retrieves the top 3 relevant chunks from the collection.

   - The retrieved chunks are combined to form a context.

   - The context and question are sent to an OpenAI model (gpt-4o-mini) to generate an answer.

3. Shutdown Process:

   - The ChromaDB collection is deleted to clean up the database.

## Requirements

To run this project, install the following dependencies:

```bash

pip install -r requirements.txt

```

The `requirements.txt` file includes:

- **FastAPI**: Framework for building the API.

- **uvicorn**: ASGI server for FastAPI.

- **python-dotenv**: For managing environment variables.

- **langchain**: Library for interacting with LLMs.

- **langchain and langchain-openai**: For interacting with OpenAI models.

- **chromadb**: Vector database for semantic searches.

- **PyPDF2**: Library for reading PDFs.

- **pydantic**: For data validation.

## How to Use

1. Set Up Your API Key: 

   - Create a .env file or set the environment variable OPENAI_API_KEY with your OpenAI API key. 

   - If not set, the application will prompt you to enter the key at startup.

2. Select a PDF: 

   - Choose a PDF of your choice and reference the path or use the example bellow.

3. Start the application by running:

   ```bash

   uvicorn main:app --reload

   ```

4. Ask a Question:

   - Send a POST request to the /ask_ai endpoint with a JSON payload like:

   ```json

   {

     "question": "Your question about the PDF content"

   }

   ```

   The application will return an answer based on the extracted content from the PDF.