Product Performance Analysis System
1. Introduction
1.1 System Overview
This Flask-based product analysis API evaluates product performance using clustering and scoring algorithms. It processes sales data to:
- Calculate key performance metrics (revenue, stock efficiency, sales velocity)
- Assign performance scores (0-100 scale)
- Classify products into performance tiers (Low/Medium/High)
1.2 Key Features
- Automated feature engineering
- MinMax normalization for fair comparison
- KMeans clustering for performance tiering
- Weighted scoring system
- Model persistence with Joblib
2. Technical Implementation
2.1 Data Flow
graph TD
A[Raw Product Data] --> B[Feature Engineering]
B --> C[Normalization]
C --> D[Scoring]
D --> E[Clustering]
E --> F[API Response]
2.2 Core Algorithms
| Component | Technology | Purpose |
|---|---|---|
| Feature Engineering | Pandas | Calculate revenue, stock ratio, sales velocity |
| Normalization | MinMaxScaler | Scale features to [0,1] range |
| Clustering | KMeans (k=3) | Performance tier classification |
3. API Documentation
3.1 Endpoint
POST /product-analysis
3.2 Request Format
[
{
"product_id": 101,
"name": "Premium Headphones",
"price": 199.99,
"units_sold": 150,
"stock": 50
},
...
]
3.3 Response Format
{
"status": "success",
"data": [
{
"product_id": 101,
"name": "Premium Headphones",
"performance_score": 85.2,
"performance_tier": "High",
"revenue": 29998.50,
...
}
],
"summary": {
"avg_score": 72.4,
"top_product": "Premium Headphones",
"total_revenue": 125499.00
}
}
4. Scoring Methodology
4.1 Feature Weights
- Revenue (50%): price × units_sold
- Stock Ratio (30%): units_sold / (stock + units_sold)
- Sales Velocity (20%): units_sold / price
4.2 Performance Tiers
- High: Top 20% of products
- Medium: Middle 60%
- Low: Bottom 20%
5. Business Applications
- Inventory optimization decisions
- Product portfolio analysis
- Marketing campaign targeting
- Supplier performance evaluation
6. Deployment
# Sample Dockerfile FROM python:3.9 RUN pip install flask pandas scikit-learn COPY . /app WORKDIR /app CMD ["flask", "run", "--host=0.0.0.0"]
7. Future Enhancements
- Time-series analysis for trend detection
- Integration with visualization dashboards
- Automated report generation
