Data Lake Storage Architecture

import matplotlib.pyplot as plt
import squarify
import numpy as np

# Data lake layers (in TB)
labels = ['Bronze Raw', 'Bronze Logs', 'Silver Cleaned', 'Silver Enriched',
          'Gold Analytics', 'Gold ML Features', 'Archive Cold', 'Staging']
sizes = [85, 120, 45, 65, 28, 35, 180, 22]
total = sum(sizes)

# Color scheme: bronze/silver/gold themed with CLAUDE.md
colors = ['#C82909', '#F54927', '#27D3F5', '#276CF5', 
          '#F5D327', '#F5B027', '#5314E6', '#6CF527']

fig, ax = plt.subplots(figsize=(12, 8), facecolor='#0a0a0f')
ax.set_facecolor('#0a0a0f')

pct = [s/total*100 for s in sizes]
squarify.plot(sizes=sizes, 
              label=[f'{l}\n{s}TB ({p:.1f}%)' for l, s, p in zip(labels, sizes, pct)],
              color=colors, alpha=0.85, ax=ax,
              text_kwargs={'fontsize': 10, 'color': 'white', 'fontweight': 'bold'})

ax.axis('off')
ax.set_title(f'Data Lake Architecture - {total/1000:.2f} PB Total', 
             fontsize=18, color='#f8fafc', fontweight='bold', pad=20)

plt.tight_layout()
plt.show()