SmartHeal - Document Q&A Chatbot 🏥💬

A sophisticated document-based question-answering chatbot that allows users to interact with their document knowledge base using Retrieval-Augmented Generation (RAG). Built with Flask, Pinecone, Cohere, and powered by advanced embeddings.

🌟 Features

• Document Embedding & Storage: Automatically processes PDF documents and stores embeddings in Pinecone
• Intelligent Chunking: Uses hierarchical document splitting for optimal retrieval
• RAG-Powered Responses: Combines document retrieval with Cohere's language model for accurate answers
• Real-time Chat Interface: Clean, responsive web interface for seamless conversations
• Session Management: Persistent chat history across page refreshes
• Audio Inpuy: Audio input avaiable for English, Hindi and Bengali.

🛠️ Technology Stack

• Backend: Flask (Python)
• Vector Database: Pinecone
• Embeddings: Cohere
• Language Model: Cohere
• Document Processing: LangChain + PyPDF
• Frontend: Bootstrap 5 + Vanilla JavaScript
• Speech-to-Text: OpenAI Whisper
• Speaker Diarization: Pyannote Audio
• Audio Processing: FFmpeg

📋 Prerequisites

• Python 3.9+
• Pinecone account and API key
• Cohere account and API key
• HuggingFace account and token
• FFmpeg (for audio conversion)

🚀 Quick Start

1. Clone the Repository

git clone https://github.com/cr7ritesh/smartheal-chatbot.git
cd Runverve

2. Install Dependencies

pip install -r requirements.txt

3. Install FFmpeg

Windows:

🎬 Install FFmpeg on Windows by TechwithMonir

macOS:

brew install ffmpeg

Ubuntu/Debian:

sudo apt update
sudo apt install ffmpeg

3. Environment Setup

Create a .env file in the root directory:

PINECONE_API_KEY=your_pinecone_api_key_here
COHERE_API_KEY=your_cohere_api_key_here
HUGGINGFACE_TOKEN=your_huggingface_token_here

4. Prepare Your Documents

• Create a docs folder in the project root
• Place your PDF files in the docs folder

5. Process Documents & Create Embeddings

python store_embed.py

This will:

• Load all PDFs from the folder
• Create intelligent chunks using hierarchical splitting
• Generate embeddings using Cohere
• Store everything in Pinecone

6. Launch the Web Application

python app.py

Visit http://localhost:5000 to start chatting with your documents!

🔧 Usage

Processing New Documents

1. Add PDF files to the documents folder
2. Run python store_embed.py to update embeddings
3. The web app will automatically use the updated knowledge base

Chatting with Documents

1. Open the web interface at http://localhost:5000
2. Ensure the system shows "Documents loaded and ready"
3. Type your questions in the chat interface
4. Get AI-powered responses based on your document content