Skip to content

himanshu-jadhav108/Smart-Inventory-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🧠 Smart Inventory AI

AI-Powered Demand Forecasting & Inventory Optimization for Retailers


🌟 Executive Summary

Retailers face a persistent challenge: balancing inventory.

  • Overstocking ties up capital and increases storage costs.
  • Stockouts lead to lost sales, frustrated customers, and damaged reputation.
  • Manual forecasting (Excel sheets, gut-feel decisions) is error-prone and reactive.

Smart Inventory AI Pro solves this problem with a modular, explainable AI system that delivers:

  • 📈 Accurate demand forecasts
  • 📦 Optimized reorder points and safety stock
  • 🔍 Anomaly detection for unusual sales patterns
  • 💡 Business-friendly insights that drive smarter decisions

The impact: higher profitability, reduced waste, and improved customer satisfaction — all in a lightweight, plug-and-play solution.


🛒 The Retail Inventory Problem

Retail shop owners often struggle with:

  • Overstocking → Capital locked in unsold goods
  • Stockouts → Lost revenue and customer churn
  • Manual Forecasting → Time-consuming, error-prone spreadsheets
  • Complex ERP Systems → Expensive, hard to implement, overkill for small shops
  • Basic Dashboards → Show past sales but lack predictive intelligence

🚀 How Smart Inventory AI Pro is Different

Existing Solutions Limitations Smart Inventory AI Pro Advantage
Excel sheets Manual, error-prone Automated AI forecasting
ERP systems Costly, complex Lightweight, modular, easy to deploy
Sales dashboards Historical only Predictive + prescriptive insights

Key Differentiators:

  • 🧠 AI-Powered Forecasting (ensemble methods for robust accuracy)
  • 💡 Explainable Insights (business-friendly recommendations)
  • Lightweight Deployment (Streamlit-based, runs locally or in cloud)
  • 📂 Plug-and-Play CSV Upload (no complex integrations required)
  • 🧩 Modular Architecture (easy to extend, customize, and scale)

💼 Why This Matters

  • Profitability Boost: Reduce capital lock-in and avoid costly stockouts.
  • Operational Efficiency: Automate forecasting and inventory planning.
  • Accessibility: Designed for small and medium retailers, not just enterprises.
  • Scalability: Works for single shops or multi-store chains.

🏗️ System Architecture

Smart Inventory AI Pro follows a modular, layered design:

  • Frontend (Streamlit) → Interactive dashboards and visualizations
  • Models Layer → Forecasting, anomaly detection, inventory optimization
  • Explainability Layer → Business insights and recommendations
  • Utilities Layer → Data loaders, helpers, validation
  • Config Layer → Themes, styling, customization

This separation of concerns ensures maintainability, extensibility, and clean interfaces.


🤖 AI & ML Techniques

  • Ensemble Forecasting: EWMA, Linear Regression, Seasonal Decomposition
  • Anomaly Detection: Statistical + IQR-based methods with severity scoring
  • Safety Stock Optimization: Dynamic calculations based on demand volatility
  • Reorder Point Logic: Lead time + demand variability for smarter replenishment

🎯 For Retailers (Simple Value Proposition)

  • No more guessing stock levels.
  • No more losing sales because of empty shelves.
  • No more wasting money on excess inventory.

👉 Just upload your sales CSV, and let Smart Inventory AI Pro tell you what to order, when, and how much.


📂 Dataset Usage

To make testing and evaluation easier, a ready-to-use kaggle dataset is included in the dataset/ folder.

This dataset allows you to:

  • Instantly explore forecasting features
  • Test anomaly detection
  • Validate inventory optimization logic
  • Understand dashboard visualizations

You can either:

  1. Use the built-in AI-generated sample data inside the app
  2. Upload the provided dataset from the dataset/ folder
  3. Upload your own retail sales CSV file

📑 Expected CSV Format

date,product_id,units_sold
2024-01-01,PROD001,45
2024-01-02,PROD002,52
2024-01-03,PROD003,48

📊 Project Structure

smart_inventory_ai/
│
├── app.py                  # Main application entry point
├── requirements.txt        # Python dependencies
├── README.md               # This file
│
├── config/
│   └── theme.py            # CSS styling and theme configuration
│
├── data/
│   └── sample_generator.py # Sample data generation
│
├── models/
│   ├── forecasting.py      # Forecasting engine with ensemble methods
│   ├── anomaly.py          # Anomaly detection algorithms
│   ├── inventory.py        # Inventory optimization logic
│   └── explainability.py   # Business insights generation
│
├── dashboard/
│   ├── charts.py           # Plotly chart generation
│   ├── metrics.py          # Metric card components
│   └── layout.py           # UI layout and components
│
├── utils/
│   ├── data_loader.py      # Data loading and validation
│   └── helpers.py          # Helper utility functions
│
└── assets/
    ├── sia_logo.jpeg       # SIA Logo
    └── styles.css          # Additional CSS styles

🖥️ Live Demo

🔗 https://smart-inventory-ai.streamlit.app/


🎬 Project Demo Video

Watch the complete walkthrough of SIA – Smart Inventory AI in action:

🔗 Click here to watch the demo video

The demo covers:

  • Problem statement
  • System architecture flow
  • Live web application walkthrough
  • Forecasting & inventory optimization
  • Explainable AI insights
  • Future vision

🔮 Future Roadmap

  • 📡 API Integration with POS systems
  • 📱 Mobile App for shop owners
  • 🧾 Multi-Product Forecasting with portfolio optimization
  • 🌍 Cloud Deployment for scalability
  • 🛒 Retail-Specific Modules (perishables, seasonal goods, promotions)

🛠️ Quick Start

# Install dependencies
pip install -r requirements.txt

# Run the application
streamlit run app.py

Upload your CSV (date, product_id, units_sold) and start optimizing inventory instantly.


📈 Model Performance

Validated with walk-forward testing:

  • MAE → Average prediction error in units
  • MAPE → Percentage-based accuracy measure

Typical accuracy:

  • <10% → Excellent

  • 10–20% → Good

  • 20% → Fair


🤝 Contributing

We welcome contributions, feature requests, and collaborations.
Open an issue or submit a pull request to join the project.


📧 Support

For questions or support, please open an issue in this repository.


Built with ❤️ by Team Eureka Fourge

Team Information

  • Himanshu Jadhav (Team Leader)
  • Yash Bhongale
  • Ritesh Gaike
  • Onkar Kharat

About

AI-powered retail inventory forecasting and optimization platform with ensemble modeling, anomaly detection, and explainable business insights.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors