LogiChat: RAG-Based Chatbot with Document Uploads

LogiChat is a Retrieval-Augmented Generation (RAG) based chatbot application powered by Large Language Models (LLMs). It enables users to upload their documents and have intelligent, context-aware conversations derived exclusively from their provided data.

Key Features

• Document Uploads: Users can upload documents in various formats.
• Intelligent Conversations: The chatbot generates contextually accurate responses based on user-provided data.
• Powered by LLMs: Leverages advanced LLMs for natural language understanding and generation.
• Secure and Private: Your documents are used solely for generating responses and are stored securely.(Not done to store the responses)

Tech Stack

• Backend: Flask, LangChain, PineConeVector, OpenAI API
• Frontend: Next.js
• Database: PineConeVector for vectorized document storage

Architecture

LogiChat uses Retrieval-Augmented Generation (RAG) to combine LLM capabilities with document retrieval. The steps include:

1. Document Vectorization: Uploaded documents are converted into embeddings and stored in a vector database.
2. History-Aware Retrieval: Queries are reformulated contextually using chat history.
3. Response Generation: Retrieved data and user queries are processed by the LLM to generate responses.

Installation and Setup

Prerequisites

• Python 3.8+
• Node.js 16+
• An OpenAI API key
• A .env file with the following keys:

  OPENAI_API_KEY=your_openai_api_key

  PINECONE_API_KEY=your_pine_cone_vector_api_key

  PINECONE_INDEX_NAME=your_pine_cone_vector_index_name

Backend Setup

1. Clone the repository:

   git clone https://github.com/naitikmp/LogiChat.git
   cd LogiChat/backend

2. Create a virtual environment and install dependencies:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt

3. Run the Flask backend:

   python app.py

client Setup

1. Navigate to the client directory:

   cd ../client

2. Install dependencies:

   npm install

3. Start the development server:

   npm run dev

Usage

1. Start the backend and client servers.
2. Open your browser and navigate to the Next.js client (typically at http://localhost:3000).
3. Upload a document to begin a conversation.
4. Interact with the chatbot using natural language queries.

File Structure

LogiChat/
├── backend/
│   ├── utils.py               # Utility files for the application
│   ├── app.py                 # Main Flask application
│   ├── requirements.txt       # Backend dependencies
│   └── .env                   # Environment variables
├── client/
│   ├── pages/                 # Next.js pages
│   ├── components/            # Reusable UI components
│   ├── public/                # Static assets
│   └── package.json           # Frontend dependencies
└── README.md                  # Project documentation

Contributing

Contributions are welcome! Please fork the repository and create a pull request with your proposed changes.