Convert plain English queries into SQL statements using a locally hosted LLM and visualize results instantly
AskYourDB is a smart web app that lets you talk to your database using natural language.
Imagine asking:
"Show me all users who signed up last month."
And the app automatically converts that into a proper SQL query, runs it, and even visualizes the result using interactive charts.
No SQL experience? No problem.
No internet? Still works — everything runs locally on your system.
This app combines:
- A locally hosted LLM (Code Llama) for converting natural language to SQL
- A Flask backend for query handling and model communication
- A dynamic frontend UI to interact with databases and visualize outputs using PyGWalker
It's designed to run entirely offline, using the GGUF format for optimized LLM inference on CPUs.
- Natural Language → SQL conversion via LLM
- Multiple Database Selection
- Table Selection UI
- Editable SQL Queries
- Instant Data Visualization with PyGWalker
- Fully Local Deployment (no cloud dependencies)
- Model Swappable (just change one line to try another LLM)
| Layer | Tech Used | Description |
|---|---|---|
| LLM | Code Llama 7B (GGUF - 5bit) | Converts text to SQL locally |
| Backend | Flask | REST API, model handling |
| Visualization | PyGWalker | Turns SQL results into charts/tables |
| Model Serving | ctransformers | Lightweight LLM runner on CPU |
| UI | HTML, CSS, JS | Clean frontend with SQL editor |
| Env Config | Python dotenv | Manage credentials and paths |
| Database | MySQl,SQLite,PostgreSQl | Manage Relational Database |
# 1. Clone the repo
git clone https://github.com/raghujhts13/text-to-sql.git
cd text-to-sql
# 2. Install dependencies
pip install -r requirements.txt
# 3. Set up environment variables (in config/.env)
# FLASK_KEY = any random string
# MODEL_PATH = path to the downloaded GGUF model
# 4. Run the app
python app.pyVisit the app at: http://127.0.0.1:5000
This project turns a complex task like database querying into something as easy as chatting. It’s lightweight, fully offline, and customizable — ideal for analysts, devs, or anyone tired of writing boilerplate SQL.
MIT License – free to use, modify, and distribute.
Pull requests, suggestions, and issues are always welcome! Feel free to fork and build your own enhanced version.