bench get-app https://github.com/raisulislam0/sales_prediction

[![Listed on Frappe Gems](https://frappegems.com/api/method/frappe_gems.seo.badge?app=raisulislam0%2Fsales_prediction)](https://frappegems.com/gems/apps/raisulislam0/sales_prediction)

📈 Sales Prediction

An AI-powered sales forecasting and predictive analytics application built natively for the Frappe/ERPNext framework.

This application leverages Meta's Prophet time-series forecasting model to analyze historical sales invoice data and generate future revenue projections directly within your ERP environment. It enables organizations to make data-driven decisions using statistical forecasting rather than relying solely on traditional sales pipeline tracking.

---

✨ Features

📊 Statistical Time-Series Forecasting

• Uses Meta Prophet to analyze historical sales trends.
• Detects seasonality, growth patterns, and recurring revenue behavior.
• Generates future sales projections with confidence intervals.

📈 Interactive Visualizations

• Powered by Plotly.js for responsive and interactive charts.
• Hover tooltips for detailed data inspection.
• Displays forecasted values alongside historical sales data.
• Includes explicit 80% confidence interval bands.

⚙️ Dynamic Aggregation Controls

• Switch between multiple forecasting frequencies:
  • Daily (D)
  • Weekly (W)
  • Monthly (MS)
• Adjust forecast horizons directly from the user interface.

🔒 Leakage-Proof Forecasting

• Automatically excludes records dated in the future.
• Prevents data leakage during model training and evaluation.
• Ensures more realistic and trustworthy forecasts.

📋 Data Insights Dashboard

• Summary cards showing:
  • Total records analyzed
  • Historical date range
  • Forecast horizon
  • Predicted revenue growth percentage
• Instant statistical insights for management and planning.

---

📋 Prerequisites

The application depends on external scientific computing libraries that must be installed inside your Bench virtual environment.

cd $PATH_TO_YOUR_BENCH

./env/bin/pip install pandas prophet

---

🚀 Installation

1. Fetch the App

bench get-app https://github.com/raisulislam0/sales_prediction.git

2. Install on Your Site

bench --site [your-site-name] install-app sales_prediction

3. Apply Migrations and Build Assets

bench --site [your-site-name] migrate
bench build

---

🖥️ Usage

1. Navigate to Sales Prediction from the ERPNext Desk.
2. Select:
   • Aggregation Frequency (Daily, Weekly, Monthly)
   • Forecast Horizon
3. Click Generate Forecast.
4. Review:
   • Historical sales trends
   • Forecasted revenue
   • Confidence intervals
   • Growth projections

---

📦 Technology Stack

• Frappe Framework v16
• ERPNext
• Python
• Pandas
• Prophet (Meta)
• Plotly.js

---

🎯 Use Cases

• Revenue forecasting
• Budget planning
• Inventory planning
• Seasonal demand analysis
• Business growth tracking
• Executive reporting

---

⚠️ Disclaimer

Forecasts are generated using historical sales data and statistical modeling techniques. Actual future sales may vary due to market conditions, operational changes, seasonal factors, and other external influences.

Predictions should be used as decision-support tools and not as guaranteed business outcomes.

---

📄 License

MIT License