🧠 Natural Language SQL AI Instantly query your internal databases using plain English — secure, offline, and schema-aware.

🚀 Overview This AI-powered tool allows non-technical users to interact with internal SQL databases using natural language. No SQL knowledge required — just ask questions like:

“List all complaints filed against John Doe.” “Show employees who accessed the R&D lab after 8 PM this month.”

Built for HR teams, police departments, IT ops, and more, the tool runs fully offline to ensure complete data privacy.

🧰 Tech Stack Python

Natural-SQL-7B (open-source LLM)

LangChain

FAISS

RAG (Retrieval-Augmented Generation)

Pandas, NumPy

Jupyter Notebooks

🔐 Key Features ✅ Natural language to SQL – Just ask questions in plain English

✅ Schema-aware – Custom-tuned to your database structure

✅ Runs locally – No data ever leaves your system

✅ Works across domains – HR, law enforcement, manufacturing, logistics, etc.

✅ Easy integration – Minimal setup, fast deployment

🧪 Example Use Cases 👮‍♂ Police Department “Show FIRs filed in Mumbai in 2024.”

“Who is assigned to Zone 3 currently?”

🏢 Smart Enterprise “List employees with no access activity in the last 10 days.”

“Who accessed the server room after working hours?”

🛠 Setup Instructions Clone the repository

Install dependencies:

bash Copy Edit pip install -r requirements.txt Launch Jupyter Notebook or your preferred IDE

Load your database schema and start querying!

📈 Impact 🔍 Improved internal data accessibility by 70%

🛠 Reduced technical support dependency by 50%

🧠 Handles over 1,000 queries/day in real-world settings