# ===================================================================
# MODULE 9, LESSON 6: CHANGE MANAGEMENT AND BUSINESS ADOPTION
# ===================================================================

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')

# Set style
sns.set_style("whitegrid")
np.random.seed(42)

print("="*70)
print("CAPSTONE PROJECT – CHANGE MANAGEMENT AND BUSINESS ADOPTION")
print("="*70)

# ----------------------------------------------------------------
# PART A: STAKEHOLDER ENGAGEMENT PLAN
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART A: Stakeholder Engagement Plan")
print("-"*60)

engagement_plan = pd.DataFrame({
    'Stakeholder': [
        'CEO / Executive Sponsor',
        'Head of Retail Banking',
        'Chief Risk Officer',
        'Loan Officers',
        'Branch Managers',
        'IT Team',
        'Compliance Officer',
        'Customers',
        'Regulators'
    ],
    'Engagement Frequency': [
        'Monthly',
        'Bi-Weekly',
        'Weekly',
        'Daily (during rollout)',
        'Weekly',
        'Weekly',
        'Monthly',
        'Ongoing',
        'Quarterly'
    ],
    'Format': [
        'Executive Steering Committee',
        'Project Review Meeting',
        'Model Validation Review',
        'Stand-ups, Training Sessions',
        'Branch Update Meetings',
        'Technical Working Group',
        'Compliance Review',
        'Communication Campaign',
        'Regulatory Reporting'
    ],
    'Key Messages': [
        'Strategic alignment, ROI, risks',
        'Progress, business impact, adoption',
        'Model performance, risk metrics',
        'How to use, benefits, support',
        'Branch performance, customer feedback',
        'Integration, security, performance',
        'Regulatory compliance, fairness',
        'Faster approvals, better service',
        'Model governance, transparency'
    ]
})

print("Stakeholder Engagement Plan:")
print(engagement_plan.to_string(index=False))

# ----------------------------------------------------------------
# PART B: TRAINING PLAN
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART B: Training Plan")
print("-"*60)

training_plan = pd.DataFrame({
    'Training Module': [
        'Module 1: Introduction to AI Lending',
        'Module 2: Using the Loan Decision API',
        'Module 3: Interpreting Model Outputs',
        'Module 4: Handling Edge Cases',
        'Module 5: Data Privacy and Security',
        'Module 6: Reporting and Monitoring'
    ],
    'Audience': [
        'All Loan Officers',
        'Loan Officers, Branch Managers',
        'Loan Officers, Risk Team',
        'Advanced Users',
        'All Employees',
        'Managers, Risk Team'
    ],
    'Duration': [
        '1 hour',
        '2 hours',
        '1.5 hours',
        '1 hour',
        '0.5 hours',
        '1 hour'
    ],
    'Format': [
        'E-learning',
        'Hands-on Workshop',
        'Workshop + Case Studies',
        'Workshop',
        'E-learning',
        'Workshop'
    ],
    'Timeline': [
        'Pre-rollout (Week 1)',
        'Pre-rollout (Week 2)',
        'Pre-rollout (Week 2)',
        'Post-rollout (Week 4)',
        'Pre-rollout (Week 1)',
        'Post-rollout (Week 4)'
    ]
})

print("Training Plan:")
print(training_plan.to_string(index=False))

# ----------------------------------------------------------------
# PART C: ADOPTION METRICS AND KPIS
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART C: Adoption Metrics and KPIs")
print("-"*60)

adoption_metrics = pd.DataFrame({
    'Metric': [
        'API Usage Rate',
        'Loan Officers Trained',
        'User Satisfaction (NPS)',
        'Automation Rate',
        'Decision Time Reduction',
        'Support Tickets',
        'System Uptime',
        'Customer NPS'
    ],
    'Target': [
        '100% of loan officers',
        '100% trained',
        '> 40',
        '> 80% of applications',
        '> 80% reduction',
        '< 10 per month',
        '99.9%',
        '> 60'
    ],
    'Measurement Frequency': [
        'Daily',
        'Monthly',
        'Monthly',
        'Daily',
        'Weekly',
        'Weekly',
        'Daily',
        'Monthly'
    ],
    'Owner': [
        'IT Ops',
        'Training Team',
        'Change Management',
        'IT Ops',
        'IT Ops',
        'IT Ops',
        'IT Ops',
        'Customer Experience'
    ]
})

print("Adoption Metrics and KPIs:")
print(adoption_metrics.to_string(index=False))

# ----------------------------------------------------------------
# PART D: SIMULATED ADOPTION DATA
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART D: Simulated Adoption Data Over Time")
print("-"*60)

def generate_adoption_data(n_weeks=26):
    """Generate simulated adoption data."""
    weeks = np.arange(1, n_weeks+1)
    
    # S-curve adoption (typical technology adoption)
    def s_curve(t, max_adoption=0.95, midpoint=12, steepness=0.4):
        return max_adoption / (1 + np.exp(-steepness * (t - midpoint)))
    
    usage_rate = s_curve(weeks, max_adoption=0.92, midpoint=10, steepness=0.35)
    # Add some noise
    usage_rate = usage_rate + np.random.normal(0, 0.02, n_weeks)
    usage_rate = usage_rate.clip(0, 0.95)
    
    # Satisfaction (NPS) improves over time
    nps = 20 + 35 * s_curve(weeks, max_adoption=0.90, midpoint=12, steepness=0.3)
    nps = nps + np.random.normal(0, 3, n_weeks)
    nps = nps.clip(0, 100)
    
    # Automation rate
    automation = 0.5 * usage_rate + 0.4 * s_curve(weeks, max_adoption=0.95, midpoint=14, steepness=0.3)
    automation = automation + np.random.normal(0, 0.02, n_weeks)
    automation = automation.clip(0, 0.90)
    
    # Support tickets (high initially, then decreasing)
    tickets = np.exp(-0.15 * weeks) * 50 + np.random.poisson(3, n_weeks)
    tickets = tickets.clip(0, 60).astype(int)
    
    # Decision time (minutes) – decreases over time
    decision_time = 60 * np.exp(-0.12 * weeks) + 5 + np.random.normal(0, 2, n_weeks)
    decision_time = decision_time.clip(5, 60)
    
    return pd.DataFrame({
        'Week': weeks,
        'Date': [datetime(2024, 1, 1) + timedelta(days=7*w) for w in range(n_weeks)],
        'Usage Rate (%)': usage_rate * 100,
        'NPS': nps,
        'Automation Rate (%)': automation * 100,
        'Support Tickets': tickets,
        'Decision Time (min)': decision_time
    })

adoption_df = generate_adoption_data(26)

# Visualise adoption data
fig, axes = plt.subplots(2, 3, figsize=(15, 10))

# 1. Usage Rate
ax = axes[0, 0]
ax.plot(adoption_df['Week'], adoption_df['Usage Rate (%)'], 'b-', linewidth=2)
ax.axhline(y=80, color='green', linestyle='--', label='Target (80%)')
ax.set_xlabel('Week')
ax.set_ylabel('Usage Rate (%)')
ax.set_title('Loan Officer Usage Rate')
ax.legend()
ax.grid(True, alpha=0.3)

# 2. NPS
ax = axes[0, 1]
ax.plot(adoption_df['Week'], adoption_df['NPS'], 'g-', linewidth=2)
ax.axhline(y=40, color='green', linestyle='--', label='Target (40)')
ax.set_xlabel('Week')
ax.set_ylabel('NPS')
ax.set_title('User Satisfaction (NPS)')
ax.legend()
ax.grid(True, alpha=0.3)

# 3. Automation Rate
ax = axes[0, 2]
ax.plot(adoption_df['Week'], adoption_df['Automation Rate (%)'], 'orange', linewidth=2)
ax.axhline(y=80, color='green', linestyle='--', label='Target (80%)')
ax.set_xlabel('Week')
ax.set_ylabel('Automation Rate (%)')
ax.set_title('Loan Automation Rate')
ax.legend()
ax.grid(True, alpha=0.3)

# 4. Support Tickets
ax = axes[1, 0]
ax.bar(adoption_df['Week'], adoption_df['Support Tickets'], color='red', alpha=0.7)
ax.set_xlabel('Week')
ax.set_ylabel('Tickets')
ax.set_title('Support Tickets (Weekly)')
ax.grid(True, alpha=0.3)

# 5. Decision Time
ax = axes[1, 1]
ax.plot(adoption_df['Week'], adoption_df['Decision Time (min)'], 'purple', linewidth=2)
ax.axhline(y=10, color='green', linestyle='--', label='Target (10 min)')
ax.set_xlabel('Week')
ax.set_ylabel('Decision Time (min)')
ax.set_title('Average Loan Decision Time')
ax.legend()
ax.grid(True, alpha=0.3)

# 6. Combined Adoption Score
ax = axes[1, 2]
adoption_score = (adoption_df['Usage Rate (%)'] * 0.3 + 
                  adoption_df['Automation Rate (%)'] * 0.3 + 
                  (adoption_df['NPS'] / 100) * 100 * 0.4)
ax.plot(adoption_df['Week'], adoption_score, 'teal', linewidth=2)
ax.axhline(y=80, color='green', linestyle='--', label='Target (80%)')
ax.set_xlabel('Week')
ax.set_ylabel('Adoption Score')
ax.set_title('Overall Adoption Score')
ax.legend()
ax.grid(True, alpha=0.3)

plt.tight_layout()
plt.savefig('adoption_dashboard.png', dpi=300, bbox_inches='tight')
plt.show()
print("Adoption dashboard saved as 'adoption_dashboard.png'")

# ----------------------------------------------------------------
# PART E: ADDRESSING RESISTANCE
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART E: Addressing Resistance – Common Concerns and Responses")
print("-"*60)

resistance = pd.DataFrame({
    'Concern': [
        "This will replace my job.",
        "The model is a black box – I can't trust it.",
        "What if the model makes a mistake?",
        "I don't have time to learn a new system.",
        "The system will take longer than my current process.",
        "What about customers who prefer human interaction?",
        "What about data privacy and security?"
    ],
    'Type': [
        'Fear of Job Loss',
        'Lack of Trust',
        'Fear of Risk',
        'Resistance to Change',
        'Perceived Inefficiency',
        'Customer Concerns',
        'Security Concerns'
    ],
    'Response': [
        "This tool augments your expertise, making you more efficient and effective.",
        "The model provides explanations (SHAP) for every decision. You can review and override.",
        "The model is validated and monitored. You have the ability to override decisions.",
        "Training is provided, and we have support available. The new system will save you time.",
        "Initial learning is temporary; the new system will reduce time by 80%.",
        "The system provides faster decisions for routine loans; complex cases still involve human interaction.",
        "Data is encrypted and access is controlled. We comply with all privacy regulations."
    ]
})

print("Resistance Management:")
print(resistance.to_string(index=False))

# ----------------------------------------------------------------
# PART F: FEEDBACK LOOP AND CONTINUOUS IMPROVEMENT
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART F: Feedback Loop and Continuous Improvement")
print("-"*60)

feedback_loop = """
--- FEEDBACK LOOP ---

1. COLLECT FEEDBACK
   - User surveys (monthly)
   - Support tickets (weekly)
   - Model performance (daily)
   - Business metrics (monthly)
   - Customer feedback (ongoing)

2. ANALYSE FEEDBACK
   - Identify pain points
   - Assess model performance
   - Measure adoption metrics
   - Evaluate business impact

3. PRIORITISE IMPROVEMENTS
   - High impact, low effort: Quick wins
   - High impact, high effort: Strategic initiatives
   - Low impact: Defer or ignore

4. IMPLEMENT IMPROVEMENTS
   - Model retraining
   - UI/UX improvements
   - Training enhancements
   - Process changes

5. COMMUNICATE
   - Share successes
   - Explain changes
   - Recognise contributions

6. REPEAT
   - Continuous cycle of improvement
"""

print(feedback_loop)

# ----------------------------------------------------------------
# PART G: SUSTAINMENT PLAN
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART G: Sustainment Plan")
print("-"*60)

sustainment_plan = {
    "Phase 1: Launch (Weeks 1-4)": {
        "Activities": [
            "Rollout to pilot branches (10% of branches).",
            "Monitor usage, support tickets, and feedback.",
            "Provide intensive support (dedicated team).",
            "Collect feedback and identify issues."
        ],
        "Success Metrics": [
            "Usage rate > 50% in pilot branches.",
            "Customer NPS > 50.",
            "Support tickets < 20 per week."
        ]
    },
    "Phase 2: Expansion (Weeks 5-12)": {
        "Activities": [
            "Rollout to all branches.",
            "Monitor adoption and performance.",
            "Adjust training based on feedback.",
            "Begin regular retraining schedule."
        ],
        "Success Metrics": [
            "Usage rate > 80% across all branches.",
            "Automation rate > 60%.",
            "Decision time < 15 minutes."
        ]
    },
    "Phase 3: Maturity (Weeks 13-26)": {
        "Activities": [
            "Optimise model and processes.",
            "Establish ongoing governance.",
            "Transition to BAU support.",
            "Begin planning for Phase 2 (commercial loans)."
        ],
        "Success Metrics": [
            "Usage rate > 90%.",
            "Automation rate > 80%.",
            "Decision time < 10 minutes.",
            "NPS > 60."
        ]
    },
    "Phase 4: Leadership (Weeks 27+)": {
        "Activities": [
            "Expand to new products.",
            "Share learnings across the organisation.",
            "Innovate with new AI capabilities.",
            "Lead industry best practices."
        ],
        "Success Metrics": [
            "Industry recognition.",
            "Thought leadership.",
            "Continuous innovation."
        ]
    }
}

for phase, details in sustainment_plan.items():
    print(f"\n{phase}:")
    print("  Activities:")
    for activity in details['Activities']:
        print(f"    • {activity}")
    print("  Success Metrics:")
    for metric in details['Success Metrics']:
        print(f"    • {metric}")

# ----------------------------------------------------------------
# PART H: FINAL PRESENTATION TO STAKEHOLDERS
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART H: Final Presentation Outline")
print("-"*60)

presentation_outline = """
--- FINAL PRESENTATION TO STAKEHOLDERS ---

1. EXECUTIVE SUMMARY (5 min)
   - What we set out to achieve.
   - What we delivered.
   - Key results and impact.

2. THE PROBLEM (5 min)
   - Slow, manual, inconsistent loan processing.
   - Competitive pressure from fintechs.
   - Regulatory and risk considerations.

3. THE SOLUTION (10 min)
   - Data-driven, AI-powered lending platform.
   - Key components: data, models, automation.
   - How it works (demo).

4. MODEL PERFORMANCE (10 min)
   - AUC: 0.85, KS: 0.42.
   - Validation and regulatory compliance.
   - Explainability (SHAP).

5. BUSINESS IMPACT (10 min)
   - ROI: 575% over 3 years.
   - Faster approvals (days → hours).
   - 80% automation.
   - Reduced defaults.

6. IMPLEMENTATION JOURNEY (10 min)
   - Timeline (12 months).
   - Key milestones.
   - Lessons learned.

7. ADOPTION AND CHANGE MANAGEMENT (10 min)
   - Usage rate: 92%.
   - User NPS: 65.
   - Training and support.

8. NEXT STEPS (5 min)
   - Expand to commercial loans.
   - Continuous improvement.
   - Innovation pipeline.

9. Q&A (15 min)
   - Open discussion.

TOTAL: 75-90 minutes

SLIDE COUNT: 25-30 slides
"""

print(presentation_outline)

# ----------------------------------------------------------------
# PART I: FINAL SUMMARY AND COMPLETION
# ----------------------------------------------------------------

print("\n" + "="*70)
print("PART I: Final Summary and Completion")
print("="*70)

print("""
CAPSTONE PROJECT COMPLETION CHECKLIST:

Phase 1: Assessment ✅
  - Project charter approved.
  - Stakeholder analysis completed.
  - Data inventory completed.

Phase 2: Data Foundation ✅
  - Data platforms built.
  - Data quality framework implemented.
  - Data pipelines developed.

Phase 3: Model Development ✅
  - Features engineered.
  - Models trained and validated.
  - Model selection completed.

Phase 4: Risk Analytics ✅
  - Expected Loss calculated.
  - Economic Capital calculated.
  - Stress testing completed.

Phase 5: Model Deployment ✅
  - Models serialised.
  - API developed.
  - Monitoring dashboard built.
  - Drift detection configured.

Phase 6: Change Management ✅
  - Training materials developed.
  - Adoption metrics defined.
  - Sustainment plan created.
  - Presentation prepared.

PROJECT SUCCESS:
  - Project delivered on time (12 months).
  - On budget ($4.8M vs $5.0M).
  - Model performance exceeds targets (AUC 0.85).
  - Business ROI 575% over 3 years.
  - Adoption rate 92%.
  - Customer NPS 65 (vs 45 baseline).

RECOMMENDATIONS:
  - Deploy XGBoost model for internal risk assessment.
  - Use Logistic Regression for regulatory submissions.
  - Continue monitoring and retraining.
  - Expand to commercial loans (Phase 2).
  - Explore additional AI capabilities (fraud detection, NLP).

CONGRATULATIONS ON COMPLETING THE CAPSTONE PROJECT!
""")

print("="*70)
print("END OF LESSON 6 – MODULE 9")
print("="*70)
print("END OF MODULE 9")
print("="*70)