A natural language interface for querying SQL databases using Google Gemini + LangChain + Streamlit. Ask business questions in plain English and get instant SQL queries, results, and AI-generated summaries.

"Show me all cancelled orders with customer names" "Which city has the most customers?" "What is the total revenue from delivered orders?"

The agent understands your question → writes SQL → runs it → explains the result in plain English.

User Question (Natural Language)
↓
Google Gemini LLM (Text-to-SQL)
↓
SQL Query Executed on SQLite Database
↓
Results + AI Business Summary
↓
Streamlit UI (Interactive Dashboard)

The project uses a sample e-commerce database with 4 tables:

• customers — customer details, city, signup date
• products — name, category, price, stock
• orders — order date, status, total amount
• order_items — product, quantity, unit price

git clone https://github.com/LakshayHasija/SQL-Agent.git
cd SQL-Agent

python -m venv venv
venv\Scripts\activate  # Windows

pip install -r requirements.txt

Create a .env file in the root directory:

GOOGLE_API_KEY=your_gemini_api_key_here

Get your free API key at aistudio.google.com

python database.py

streamlit run app.py

• What is the total revenue from all orders?
• Show me all customers from Delhi
• Which product has the highest price?
• How many orders were delivered vs cancelled?
• Who placed the most orders?
• List all electronics products under ₹50,000

• Natural Language to SQL — No SQL knowledge needed
• AI Business Summaries — Results explained in plain English
• Query Validation — Prevents harmful queries (DROP, DELETE etc.)
• Interactive UI — Clean Streamlit dashboard with example questions
• Modular Architecture — Easily swap database or LLM provider

sql-agent/
├── app.py          # Streamlit UI
├── agent.py        # LLM + SQL orchestration logic
├── database.py     # DB schema, sample data, query runner
├── requirements.txt
└── .gitignore

• Support for PostgreSQL / MySQL databases
• Multi-turn conversational memory
• Chart generation from query results
• Authentication for multi-user access