Website Q&A Assistant

An AI-powered Website Question & Answer Assistant built using Google Gemini, FastMCP, Python, and Streamlit.

The application allows users to provide a website URL and ask questions about its content in natural language. Website content is retrieved through a FastMCP server, processed, and then analyzed by Google's Gemini model to generate accurate responses.

---

Features

• Ask questions about any public website
• Website retrieval through FastMCP
• AI-powered answers using Google Gemini
• Streamlit-based user interface
• Deployable on Google Cloud Run
• Website content inspection for transparency

---

Architecture

User → Streamlit Frontend → FastMCP Client → FastMCP Server → scrape_website Tool → Website Content Retrieval → Google Gemini → Generated Answer

---

Technologies Used

• Google Gemini API
• FastMCP
• Python
• Streamlit
• BeautifulSoup
• Requests
• Google Cloud Run

---

Example Use Cases

• University website analysis
• Scholarship information lookup
• Technical documentation summarization
• Organization and business research
• General website content exploration

---

Project Structure

app.py

• Streamlit frontend application

mcp_server.py

• FastMCP server implementation

requirements.txt

• Python dependencies

Dockerfile

• Application containerization

Dockerfile.mcp

• MCP server containerization

---

How It Works

1. User enters a website URL.
2. User asks a question.
3. FastMCP retrieves website content.
4. Content is cleaned using BeautifulSoup.
5. Gemini analyzes the content.
6. A natural language answer is generated.

---

Author

Yelaka Roshan Kumar

GitHub: https://github.com/yroshan-dev

LinkedIn: https://www.linkedin.com/in/roshan-y