🤖 Dual Mode RAG Chatbot

Intelligent Document Q&A with Context-Aware Responses

📹 Live Demo (https://covid-rag.streamlit.app)

🎯 Quick Demo: Upload a PDF → Ask questions → Get intelligent, source-cited answers in seconds

🚀 The best part?

Two Intelligent Modes in One App:

• 🩺 COVID-19 Assistant: Pre-loaded with NIH treatment guidelines
• 🧠 Custom Document Bot: Upload your own PDFs/docs for instant Q&A

Key Features:

• ⚡ Real-time RAG: Semantic search + GPT responses with conversation memory
• 📚 Smart Citations: Every answer includes source page references
• 🧠 Context Awareness: Remembers conversation history for follow-up questions
• 📁 Multi-format Support: PDF, TXT, Markdown files
• 🎨 Clean UI: Intuitive Streamlit interface with mode switching
• 💾 Persistent Storage: ChromaDB vector database with session management

🛠️ Technical Architecture

📄 Documents → 🔄 Text Splitting → 🧮 Embeddings → 🗃️ ChromaDB
                                                        ↓
🤖 LLM Response ← 📝 Prompt Engineering ← 🔍 Semantic Search

Tech Stack:

• Backend: Python, LangChain, ChromaDB
• Embeddings: HuggingFace (all-MiniLM-L6-v2)
• LLM: OpenAI GPT-3.5-turbo via OpenRouter
• Frontend: Streamlit
• Vector Store: ChromaDB with cosine similarity

⚡ Quick Start

1. Clone & Install

git clone https://github.com/fomativeh/RAG-Chatbot.git
cd RAG-Chatbot
pip install -r requirements.txt

2. Environment Setup

Create .env file:

OPENROUTER_API_KEY=your_openrouter_api_key_here
OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
COVID_DOC_URL=https://www.ncbi.nlm.nih.gov/books/NBK570371/pdf/Bookshelf_NBK570371.pdf

3. Launch

streamlit run app.py

4. Usage Options

Option A - COVID Mode:

• Add PDF files to data/covid_docs/ directory
• Select "COVID-19 Assistant" mode
• Start asking medical guideline questions

Option B - Custom Mode:

• Select "Your Custom Bot" mode
• Upload PDFs via the sidebar
• Ask questions about your documents

🎯 Core Features Breakdown

🔍 Intelligent Retrieval

• Semantic search using HuggingFace embeddings
• Top-K document retrieval (configurable)
• Metadata preservation for accurate citations

💬 Conversation Memory

• Maintains context across multiple exchanges
• Configurable conversation history depth
• Smart prompt engineering with context injection

📊 Document Processing

• Recursive text splitting for optimal chunks
• Multi-format support (PDF, TXT, MD)
• Automatic metadata extraction (source, page numbers)

🗄️ Data Management

• Persistent ChromaDB storage
• Session-based collections for uploaded files
• Automatic cleanup and error handling

📋 Requirements

streamlit>=1.28.0
langchain>=0.1.0
langchain-openai>=0.0.5
langchain-community>=0.0.13
chromadb>=0.4.15
sentence-transformers>=2.2.2
PyPDF2>=3.0.1
python-dotenv>=1.0.0
requests>=2.31.0

🎨 UI/UX Highlights

• Mode Switching: Toggle between COVID and custom document modes
• File Upload: Drag-and-drop interface with progress indicators
• Real-time Status: Loading spinners, success/error messages
• Citation Display: Source references with page numbers
• Conversation Management: Clear chat functionality
• Responsive Design: Clean, professional Streamlit interface

🔧 Advanced Configuration

Custom Prompt Templates

Modify prompts in /prompts/ directory:

• covid_assistant_prompt.txt - COVID mode system prompt
• custom_bot_prompt.txt - Custom document mode prompt