🤖 ConvoWeb

ConvoWeb is a Conversational RAG (Retrieval-Augmented Generation) application built with Streamlit, LangChain, Google Gemini, and Chroma Vector Store. It allows users to ask questions based on the contents of a given website by loading and indexing its content, then using a history-aware conversational interface to provide accurate, context-rich responses.

---

🧠 Features

• ✅ Load and parse webpage content using WebBaseLoader
• ✅ Split content into manageable chunks with RecursiveCharacterTextSplitter
• ✅ Embed content using GoogleGenerativeAIEmbeddings
• ✅ Store and retrieve chunks with Chroma vector store
• ✅ Use LangChain's create_history_aware_retriever for contextual search
• ✅ Gemini-powered response generation via ChatGoogleGenerativeAI
• ✅ Chat memory via st.session_state for multi-turn conversations
• ✅ Simple and intuitive Streamlit UI

---

📷 Design Overview

RAG Pipeline Diagram

🎥 Demo Video

---

🚀 How to Run

🔧 Requirements

• Python 3.9+
• Packages listed in requirements.txt (see below)
• Google Gemini API Key in a .env file

🧪 Setup Instructions

# Clone the repository
git clone https://github.com/Nikhil-Dadhich/ConvoWeb.git
cd ConvoWeb

# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows

# Install dependencies
pip install -r requirements.txt

# Add your Google API key to .env
echo "GEMINI_API_KEY=your_google_gemini_key" > .env

# Run the Streamlit app
streamlit run app.py

---

🔐 Environment Variables

Create a .env file in the root directory:

GEMINI_API_KEY=your_google_gemini_api_key

---

🗣️ Example Usage

1. Enter a valid website URL in the sidebar.
2. Wait while the content is parsed and vectorized.
3. Ask questions in the chat interface based on the content.
4. Enjoy rich, contextual answers from the chatbot!

---

🧩 Components

• get_vectorstore(): Loads and splits the website content and stores in Chroma.
• get_context_retriever_chain(): Creates a retriever chain using chat history.
• get_conversational_rag_chain(): Combines retriever and LLM for RAG.
• Streamlit App: Handles user input and manages chat state.

---

👨‍💻 Author

Nikhil Dadhich

Built with ❤️ using Streamlit and LangChain.