🚀 PredictX – AI Sales Forecasting Dashboard

PredictX is a modern AI-powered sales forecasting dashboard that predicts future sales trends using Linear Regression and visualizes results with interactive charts.

The application allows users to upload historical sales data in CSV format, train a simple machine learning model, and generate forecasts with clear analytics and visual insights.

---

✨ Features

📂 Upload CSV dataset with historical sales data 📈 Predict future sales using Linear Regression 📊 Interactive data visualization using charts 📉 Model evaluation with MAE (Mean Absolute Error) 📋 Analytics dashboard for brand sales breakdown ⚡ Fast client-side processing (no backend required)

---

🧠 Machine Learning Approach

PredictX uses a Linear Regression time-trend model to forecast sales.

Workflow:

1. Upload historical sales CSV
2. Parse and validate the dataset
3. Convert dates into a numerical time index
4. Train a Linear Regression model
5. Generate predictions for future days
6. Evaluate model performance

Model metrics include:

• Mean Absolute Error (MAE)
• Approximate prediction accuracy

---

📊 Dashboard Analytics

The application provides multiple insights:

• Sales forecast visualization • Historical vs predicted comparison • Brand sales distribution • Sales share percentage by brand

Charts are powered by Recharts for smooth and interactive visualization.

---

🛠 Tech Stack

Frontend

• React
• TypeScript
• Vite

UI / Styling

• Tailwind CSS
• Radix UI
• Lucide Icons

Data Visualization

• Recharts

Routing

• React Router

Utilities

• React Hook Form
• Class Variance Authority
• Tailwind Merge

---

📂 Project Structure

predictx
│
├─ src
│   ├─ pages
│   │   └── CarDashboard.tsx
│   │
│   ├─ components
│   │   └── ui
│   │
│   ├─ App.tsx
│   ├─ main.tsx
│   └─ index.css
│
├─ package.json
├─ vite.config.ts
└─ index.html

---

⚙️ Installation

Clone the repository

git clone https://github.com/diyuworks/predictx.git

Go to the project directory

cd predictx

Install dependencies

npm install

Run the development server

npm run dev

Open in browser

http://localhost:5173

---

📄 Example CSV Format

The dataset must include:

date,sales,brand

Example:

2024-01-01,120,Toyota
2024-01-02,140,Toyota
2024-01-03,135,Honda

---

📸 Screenshots

(Add screenshots here after uploading them to the repository)

Example sections:

Dashboard

Prediction Chart

Results Panel

---

🔮 Future Improvements

• Add advanced ML models (Random Forest / ARIMA) • Deploy live web application • Export predictions as downloadable reports • Add authentication system • Improve forecasting accuracy

---

👩‍💻 Author

Diya Malviya Computer Science Student | AI & Full Stack Enthusiast

GitHub https://github.com/diyuworks

---

⭐ If you like this project, consider giving it a star!