---

🛒 Northwind Sales Chatbot:

An AI-powered chatbot that interacts with the classic Northwind Sales Database, enabling users to ask natural language questions about sales, products, customers, and orders. The chatbot translates user queries into SQL (via LLM) and returns accurate business insights in real-time.

---

✨ Features

✅ Natural Language Queries – Ask questions like "Which product sold the most in Q2?" or "Top 5 customers by revenue last year" ✅ Conversational AI – Uses LLMs to understand and respond intelligently ✅ Database Integration – Connects to the Northwind database for live data retrieval ✅ Sales Insights – Generate KPIs, trends, and visual summaries ✅ User-friendly UI – Chat interface for easy interaction ✅ Extensible – Can plug into other databases or APIs with minimal changes

---

🏗️ Tech Stack

• Backend: Python (FastAPI / Flask)
• Frontend: React + Tailwind (chat UI)
• Database: Northwind SQL (SQLite / SQL Server / PostgreSQL)
• AI Layer: LLM (OpenAI / Groq / custom adapter)
• Authentication: Optional (JWT/OAuth)
• Deployment: Docker + GitHub Actions CI/CD

---

🚀 Getting Started

1️⃣ Clone the repo

git clone https://github.com/your-username/northwind-sales-chatbot.git
cd northwind-sales-chatbot

2️⃣ Install dependencies

# Backend
cd backend
pip install -r requirements.txt

# Frontend
cd frontend
npm install

3️⃣ Setup environment

Create a .env file in the backend folder:

OPENAI_API_KEY=your_api_key_here
DATABASE_URL=sqlite:///northwind.db

4️⃣ Run locally

# Backend
uvicorn main:app --reload

# Frontend
npm run dev

App will be available at http://localhost:3000 🎉

---

💡 Example Queries

• "Show me the top 10 best-selling products"
• "Who are the top 5 customers in 1997?"
• "What was the monthly revenue trend last year?"
• "Which salesperson closed the highest number of orders?"

---

📊 Sample Output:

✅ Natural language → SQL → Insight

Query: "Top 5 customers by sales in 1997" SQL Generated:

SELECT c.CompanyName, SUM(od.Quantity * od.UnitPrice) AS TotalSales
FROM Orders o
JOIN Customers c ON o.CustomerID = c.CustomerID
JOIN [Order Details] od ON o.OrderID = od.OrderID
WHERE strftime('%Y', o.OrderDate) = '1997'
GROUP BY c.CompanyName
ORDER BY TotalSales DESC
LIMIT 5;

Result: Table + chart with top customers

---