🧠 Natural Language to SQL Interface

Query any SQLite database using plain English!
This project converts natural language to SQL using LangChain and Groq’s LLaMA3 model, executes the generated SQL against a specified SQLite database, and displays results in a sleek Streamlit web interface.

---

📁 Project Structure

.
├── app.py          # Streamlit app for user interaction
├── database.py     # Optional script to create/populate a sample DB
├── your_database.db# SQLite DB file (user-provided or auto-loaded)
├── requirements.txt
└── README.md

---

📊 Features

• 💬 Convert English questions into SQL using LLaMA3 via Groq API
• 📂 Support for any SQLite database
• 🧠 Auto-extracts and feeds database schema to the LLM
• 📊 View query results in an interactive table
• 📥 Download results as CSV
• 🔍 Toggle to view a plain-English explanation of the SQL
• 📈 Auto-render bar charts for COUNT, AVG, SUM queries

---

🔧 How to Adapt for Any Database

To support any SQLite .db file:

1. Replace the default student.db with your own .db file

2. In app.py, modify the DB_PATH variable:

   DB_PATH = "your_database.db"

3. Make sure the schema is readable. The code automatically extracts table/column names using:

   SELECT name FROM sqlite_master WHERE type='table';
   PRAGMA table_info(table_name);

4. This schema is sent to LLaMA3 to help it generate accurate SQL for your specific database.

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/yourusername/nl2sql-interface.git
cd nl2sql-interface

2. Install Dependencies

Create a requirements.txt file with:

streamlit
pandas
langchain
langchain-groq

Then install:

pip install -r requirements.txt

3. Set Groq API Key

For Linux/macOS:

export GROQ_API_KEY=your_groq_api_key

For Windows (CMD):

set GROQ_API_KEY=your_groq_api_key

4. Provide Your SQLite Database

• Use student.db (provided), or
• Replace it with your own .db file and update the DB_PATH in app.py

5. Launch the App

streamlit run app.py

---

🔍 Example Natural Language Queries

Try questions like:

• "How many records are there in each table?"
• "Show all entries from the employees table."
• "What is the average salary by department?"
• "List customers who spent more than $500."

---

🧰 Tech Stack

• Python 3.7+
• SQLite – lightweight relational database
• Streamlit – frontend UI
• LangChain – framework for LLM chaining
• Groq + LLaMA3 – LLM backend for NL→SQL

---

📄 License

This project is licensed under the MIT License.

---

🙏 Acknowledgements

Thanks to these amazing tools:

• LangChain
• Groq
• Streamlit

---

✨ Build once, query any database with the power of language + LLMs!