Co-developed with @burak-basoglu

Live Dashboard: https://hypeproject.streamlit.app/

📦 Demand Forecasting-Based Inventory Management System

A comprehensive, data-driven decision support system designed to optimize inventory management through demand forecasting, model comparison, and customer-centric insights.

This project goes beyond traditional forecasting by integrating business logic, analytics, and usability layers into a single workflow.

---

🧩 Business Problem

Inventory management is a critical challenge in retail operations.

• Understocking leads to lost sales and poor customer experience
• Overstocking results in increased holding costs and inefficient capital allocation

To enable effective planning, businesses must answer:

• What will be the future demand for each product?
• Which products require immediate replenishment?
• How can forecasting outputs be translated into actionable decisions?
• Which customer segments should be targeted under stock constraints?

This project addresses these questions by building an end-to-end forecasting and decision support system.

---

🎯 Project Objectives

• Forecast product demand using historical sales data
• Compare different forecasting approaches and evaluate performance
• Support inventory planning with actionable insights
• Aggregate demand across multiple time levels:
  • Weekly
  • Monthly
  • Quarterly
• Integrate RFM segmentation to connect inventory decisions with customer value
• Deliver outputs in a dashboard-ready format

---

🧠 Solution Architecture

The system is structured as a multi-layer analytical pipeline:

1. Data Preparation

• Cleaning transactional data
• Handling missing values
• Removing anomalies and invalid entries
• Aggregating data into time-based structures

2. Exploratory Data Analysis (EDA)

• Understanding demand distribution and patterns
• Identifying seasonality and trends
• Extracting business-relevant insights

3. Demand Forecasting

• Modeling future demand using historical data
• Generating forecasts across different time granularities
• Supporting operational planning

4. Model Evaluation

• Comparing forecasting models
• Storing results in model_comparison.csv
• Selecting the most effective model

5. Customer Segmentation (RFM)

• Identifying high-value and at-risk customers
• Supporting marketing and inventory decisions
• Linking demand insights with customer behavior

6. Application Layer (Streamlit)

• Providing an interactive interface
• Making results accessible to non-technical users
• Enabling quick business interpretation

---

⚙️ Methodology

📊 Demand Forecasting

• Time-based aggregation (weekly, monthly, quarterly)
• Forecast generation using historical demand patterns
• Model comparison and evaluation

👥 RFM Analysis

• Recency → How recently a customer purchased
• Frequency → How often they purchase
• Monetary → How much they spend

📈 Model Evaluation

• Comparative performance tracking
• Selection based on predictive effectiveness

---

🧮 Data Assets

The project includes multiple processed datasets:

Raw & Processed Data

• df.csv
• df_merged.csv

Time-Based Aggregations

• weekly_sales.csv
• monthly_sales.csv
• quarterly_sales.csv

Aggregated Outputs

• aggregated_weekly_sales.csv
• aggregated_monthly_sales.csv
• aggregated_quarterly_sales.csv

Model Evaluation

• model_comparison.csv

Dashboard Outputs

• dashboard_summary.csv
• dashboard_summary_main.csv

Customer Segmentation

• rfm_results.csv

---

🛠️ Tech Stack

Programming Language

• Python

Core Libraries

• pandas
• numpy
• scikit-learn
• tensorflow

Visualization & App Layer

• streamlit
• altair

---

🚀 How to Run

1. Clone the repository

git clone "repository-url"
cd Demand_Forecasting_Based_Inventory_Management_System

2. Install dependencies

pip install -r requirements.txt

3. Run analysis

python main.py

4. Launch Streamlit app

streamlit run app.py

📊 Business Value

This system provides measurable impact in real-world scenarios:

✅ Reduced Stockout Risk

Forecast-driven planning helps maintain product availability.

✅ Lower Inventory Costs

Prevents overstocking and reduces capital lock-in.

✅ Smarter Decision-Making

Combines forecasting with segmentation for better strategy.

✅ Cross-Functional Value

Useful for:

• Data teams
• Supply chain teams
• Marketing teams

---

💡 Example Use Cases

• Identifying products with upcoming high demand
• Prioritizing inventory replenishment
• Designing targeted campaigns for high-value customers
• Monitoring performance through dashboards

---

🔍 Key Strengths of the Project

This project demonstrates:

• Strong problem framing
• Integration of DL + business logic
• Focus on decision support, not just prediction
• Multi-level data aggregation
• Customer-centric thinking

This is not just a forecasting model —
it is a decision support system for inventory optimization.

---

🔮 Future Improvements

• Safety stock calculation
• Reorder point optimization
• Lead time integration
• API deployment
• Automated retraining pipeline
• Real-time dashboard integration