AI SQL Analytics System using MCP

An AI-powered SQL analytics system that allows users to query a MySQL database using natural language. The system converts user questions into SQL queries, executes them through a Model Context Protocol (MCP) server, and generates visualizations and insights automatically.

This project demonstrates how AI systems can interact with relational databases to enable AI-assisted data analysis workflows.

---

Live Project Demo

Interactive portfolio site:

https://mcp-mysql-site.vercel.app/

The demo showcases:

• Natural language queries • Automatic SQL generation • Database execution • Data visualization • AI-generated insights

---

Project Architecture

The system architecture connects AI, database, and analytics components.

Workflow:

User Query ↓ Claude AI ↓ MCP Server (Python) ↓ MySQL Database ↓ Data Processing ↓ Visualization ↓ AI Insight Generation

---

Key Features

Natural Language Database Queries Users ask questions in plain English instead of writing SQL.

AI Generated SQL Claude AI converts natural language questions into executable SQL queries.

MySQL Integration Queries are executed on a MySQL database using Python connectors.

Data Visualization Query results are converted into charts for easier interpretation.

AI Generated Insights The system summarizes results and highlights key patterns automatically.

Interactive Portfolio Demo A modern technology-themed interface demonstrates the full pipeline.

---

Example Queries

Examples of questions that can be asked:

Show revenue by film category

Which customers spend the most money?

Find monthly rental trends

Top rented films

Revenue comparison by store location

---

Example Visualization

Example output generated from the Sakila database.

---

Demo Screenshot

Example interaction showing AI query execution and visualization.

---

Technology Stack

Python MySQL Model Context Protocol (MCP) Claude AI Matplotlib Pandas

---

Project Structure

mcp_mysql
│
├── charts
│   └── revenue_by_category.png
│
├── demo
│   └── demo_query.png
│
├── config
│   └── claude_desktop_config_sample.json
│
├── architecture.png
├── db_mcp_server.py
├── db_tools.py
├── example_queries.sql
├── requirements.txt
└── README.md

---

Installation

Clone the repository

git clone https://github.com/Peruks/mcp_mysql.git

Navigate to the project directory

cd mcp_mysql

Install dependencies

pip install -r requirements.txt

---

Running the MCP Server

Start the MCP server:

python db_mcp_server.py

The server exposes database operations to Claude AI through the Model Context Protocol.

---

Claude Desktop Configuration

To connect Claude Desktop to the MCP server, update your Claude configuration file.

Location (Windows):

C:\Users\<username>\AppData\Roaming\Claude\claude_desktop_config.json

Example configuration:

{
  "mcpServers": {
    "mysql-tools": {
      "command": "python",
      "args": [
        "C:\\mcp_mysql_project\\db_mcp_server.py"
      ]
    }
  }
}

---

Database

This project uses the Sakila sample database, a MySQL dataset representing a DVD rental business.

Key tables include:

customer film category payment rental inventory

---

Use Cases

AI-assisted data exploration Automated SQL generation Conversational database analytics Rapid business insight generation AI-powered analytics dashboards

---

Future Improvements

Real-time dashboards Interactive chart generation Automated report generation Support for multiple databases Advanced AI insight generation

---

Author

Perarivalan KS

AI / Data Analytics enthusiast focused on building real-world AI-powered analytics systems.

Portfolio Project Demo:

https://mcp-mysql-site.vercel.app/

---