SQL Agent – Natural Language Database Interface (v2)

Using LangChain-Langgraph

v1 without Langgraph

📌 Overview

This project implements a natural-language to SQL agent that allows users to query and modify multiple mock databases using plain English. It supports CRUD operations with a human-in-the-loop verification system to ensure safe database modifications.

The project is built with:

• Python + FastAPI for backend REST API
• SQLite for multiple mock databases (HR, Healthcare, E-commerce, Finance, Education)

---

🚀 Features

• 🔹 Interact with 5 mock databases (hr.db, healthcare.db, ecommerce.db, finance.db, education.db)
• 🔹 Natural language CRUD support (Read/Write/Update/Delete)
• 🔹 Human-in-the-loop approval for Create/Update/Delete before commit
• 🔹 Pending queue manager to track and approve/reject modifications
• 🔹 FastAPI REST endpoints with interactive Swagger UI

---

⚙️ Setup

1. Clone the Repository

git clone https://github.com/your-username/mini-sql-agent.git
cd mini-sql-agent

2. Create Virtual Environment (Optional)

python -m venv venv
source venv/bin/activate   # on Linux/Mac
venv\Scripts\activate      # on Windows

3. Install Dependencies

pip install -r requirements.txt
after completed,
pip install langchain_google_genai

4. Environment Variables

Create a .env file in the project root and add your Gemini API key:

GEMINI_API_KEY=your_api_key_here

5. Run FastAPI App

with Uvicorn:

uvicorn main:app --reload --port 8000

---

🖥️ API Usage

Once running, access:

• API Docs: [http://localhost:8000/docs] (http://localhost:8000/docs) and initialize database and test your nl-queries
• Example endpoint: /query?text=Show all employees in HR&db=hr
• Make sure that the backend url set in app.js is same as localhost port, else change it accordingly.

---

FASTAPI Screenshots and also with Frontend.

LangGraph Workflow

✅ Status

This is a classic agentic made with (langchain-LangGraph) but yet slow with other LLMs. Do try LangSmith Integration, added in the requirements. Add your LangSmith api directly via env file.

---

=======