InsightSQL

An AI-powered Text-to-SQL Analytics Engine that converts natural language into SQL, executes it on a MySQL database, visualizes results with interactive dashboards, and automatically generates AI-powered business insights.

---

Overview

InsightSQL enables users to interact with relational databases using natural language instead of writing SQL manually.

Users simply ask questions such as:

"Show the top 10 states by total sales revenue."

The system automatically:

• Understands the question using Gemini 2.5 Flash
• Generates optimized SQL
• Executes the query on MySQL
• Displays the results as an interactive table
• Creates dynamic visualizations
• Produces AI-generated business insights
• Evaluates SQL quality using RAGAS

---

Key Features

AI-Powered SQL Generation

• Natural Language → SQL
• Powered by Gemini 2.5 Flash
• LangChain prompt engineering
• Automatic SQL cleaning & validation

---

Database Integration

• MySQL backend
• SQLAlchemy engine
• Automatic schema discovery
• Dynamic table explorer

---

Interactive Analytics

• Interactive Data Tables
• Plotly Visualizations
• Bar Charts
• Line Charts
• Area Charts
• Automatic chart selection

---

AI Business Insights

After executing a query, the application automatically generates:

• Executive summary
• Key trends
• Business observations
• Actionable insights

---

Evaluation Dashboard

• RAGAS Metrics
• SQL correctness
• Answer relevance
• Semantic evaluation
• LLM-as-a-Judge (Llama-3 via Groq)

---

System Architecture

                   Natural Language Query
                           │
                           ▼
                  Streamlit User Interface
                           │
                           ▼
                LangChain Prompt Pipeline
                           │
                           ▼
               Database Schema Retrieval
                           │
                           ▼
                  Gemini 2.5 Flash LLM
                           │
                           ▼
                  SQL Query Generation
                           │
                           ▼
                 SQL Validation & Cleaning
                           │
                           ▼
                 SQLAlchemy + MySQL Engine
                           │
                           ▼
                    Pandas DataFrame
                           │
          ┌────────────────┴────────────────┐
          ▼                                 ▼
   Interactive Data Table          Plotly Visualizations
          │                                 │
          └────────────────┬────────────────┘
                           ▼
                 AI Business Insights
                           │
                           ▼
                    RAGAS Evaluation

---

Application Screenshots

Home Page

---

SQL Generation & Execution

---

Interactive Dashboard

---

Evaluation Dashboard

---

Tech Stack

Category	Technology
Programming Language	Python
Frontend	Streamlit
LLM	Gemini 2.5 Flash
AI Framework	LangChain
Database	MySQL
ORM	SQLAlchemy
Data Processing	Pandas
Visualization	Plotly
Evaluation	RAGAS
Judge Model	Llama-3 (Groq)

---

Project Structure

InsightSQL
│
├── Data/
│   ├── Customers.csv
│   ├── Products.csv
│   ├── Regions.csv
│   ├── sales_order.csv
│   ├── State_Regions.csv
│   └── 2017_Budgets.csv
│
├── images/
│
├── notebooks/
│
├── app.py
├── sql_engine.py
├── requirements.txt
├── README.md
├── .env.example
└── docker-compose.yml

---

Installation

Clone Repository

git clone https://github.com/VisheshJain28/InsightSQL.git

---

Navigate to Project

cd InsightSQL

---

Install Dependencies

pip install -r requirements.txt

---

Configure Environment Variables

Create a .env file.

GOOGLE_API_KEY=YOUR_GOOGLE_API_KEY

DB_HOST=localhost
DB_PORT=3306
DB_USER=root
DB_PASSWORD=your_password
DB_NAME=your_database

---

Run the Application

streamlit run app.py

---

Example Queries

• Show the top 7 states by total sales revenue.
• Find the products with the highest sales.
• Display total revenue generated by each region.
• Which customers generated the highest revenue?
• Compare state-wise sales.
• Show the monthly sales trend.
• List all available products.
• Find the budget allocated to Product 12.

---

Future Improvements

• Query History
• Download Results as CSV/Excel
• User Authentication
• Dashboard Customization
• Conversational Memory
• Voice-to-SQL
• Multi-turn Analytics