AI SQL Generator

A conversational AI application designed to generate synthetic data and perform natural language data querying on SQL schemas. This project uses Google's Gemini 2.0 Flash model, Streamlit for the UI, and PostgreSQL for data storage.

Prerequisites

• Python: 3.13+
• uv: Fast Python package installer and resolver
• Docker & Docker Compose: For running the database and application containerized.

Installation

1. Clone the repository:

   git clone <repository-url>
   cd ai-app

2. Install uv (if not installed):

   # On macOS/Linux
   curl -LsSf https://astral.sh/uv/install.sh | sh
   
   # On Windows
   powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

3. Install Dependencies: This command creates a virtual environment and installs both project and development dependencies (like pytest, pylint, mypy).

   uv sync --extra dev

4. Environment Setup: Create a .env file in the root directory based on your configuration:

   # .env
   # Google Cloud / Vertex AI
   GOOGLE_CLOUD_PROJECT=your-project-id
   GOOGLE_CLOUD_LOCATION=us-central1
   
   # Database
   DB_USER=postgres
   DB_PASSWORD=postgres
   DB_NAME=synthetic_db
   DB_HOST=db
   DB_PORT=5432
   
   # Langfuse
   LANGFUSE_SECRET_KEY=sk-lf-...
   LANGFUSE_PUBLIC_KEY=pk-lf-...
   LANGFUSE_HOST=https://cloud.langfuse.com

Running the Application

This spins up both the Streamlit app and the PostgreSQL database.

docker compose up

Code Quality & Testing

This project uses pylint for linting and mypy for static type checking.

Run Linting:

uv run pylint src tests

Run Type Checking:

uv run mypy

Run Tests:

uv run pytest