CONCLUSION
       /     |     \
   Argument  Argument  Argument
      |        |          |
   Evidence  Evidence  Evidence

Power/Interest Matrix:

High Power, Low Interest    |  High Power, High Interest
(Keep Satisfied)            |  (Manage Closely)
                            |
----------------------------+----------------------------
Low Power, Low Interest     |  Low Power, High Interest
(Monitor)                   |  (Keep Informed)

# ===================================================================
# MODULE 8, LESSON 5: COMMUNICATION AND STAKEHOLDER MANAGEMENT
# ===================================================================

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("COMMUNICATION AND STAKEHOLDER MANAGEMENT FOR DATA SCIENTISTS")
print("="*70)

# ----------------------------------------------------------------
# PART A: EXECUTIVE SUMMARY TEMPLATE
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART A: Executive Summary Template")
print("-"*60)

def generate_executive_summary(project_name, objective, key_findings, recommendation, 
                              roi, timeline, risk_level):
    """Generate an executive summary for a data science project."""
    summary = f"""
--- EXECUTIVE SUMMARY ---
PROJECT: {project_name}
DATE: {datetime.now().strftime('%B %d, %Y')}
AUTHOR: Data Science Team

1. OBJECTIVE
{objective}

2. KEY FINDINGS
{chr(10).join(['   • ' + f for f in key_findings])}

3. RECOMMENDATION
{recommendation}

4. BUSINESS IMPACT
   • Estimated ROI: {roi}
   • Timeline: {timeline}
   • Risk Level: {risk_level}

5. NEXT STEPS
   • [Action 1]
   • [Action 2]
   • [Action 3]

6. SIGN-OFF
   Data Science Lead: _______________
   Business Sponsor: _______________
   Risk/Compliance: _______________
"""
    return summary

# Example executive summary
example_summary = generate_executive_summary(
    project_name="AI-Powered Fraud Detection System",
    objective="Reduce credit card fraud losses by implementing a real-time machine learning model.",
    key_findings=[
        "The new model detects 98% of fraudulent transactions, compared to 92% for the current system.",
        "False positives are reduced by 60%, saving $2M annually in manual review costs.",
        "The model meets all regulatory requirements for fairness and explainability."
    ],
    recommendation="Deploy the model in production and schedule weekly retraining.",
    roi="250% over 3 years ($5M savings on $2M investment)",
    timeline="6 months",
    risk_level="Medium"
)

print(example_summary)

# ----------------------------------------------------------------
# PART B: STAKEHOLDER MAPPING VISUALISATION
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART B: Stakeholder Mapping Visualisation")
print("-"*60)

def create_stakeholder_map():
    """Create a Power/Interest matrix visualisation."""
    fig, ax = plt.subplots(figsize=(10, 8))
    
    # Define quadrants
    ax.set_xlim(0, 10)
    ax.set_ylim(0, 10)
    
    # Draw quadrant lines
    ax.axhline(y=5, color='black', linestyle='--', alpha=0.3)
    ax.axvline(x=5, color='black', linestyle='--', alpha=0.3)
    
    # Quadrant labels
    ax.text(2.5, 8.5, 'High Power, Low Interest\n(Keep Satisfied)', ha='center', fontsize=10, style='italic')
    ax.text(7.5, 8.5, 'High Power, High Interest\n(Manage Closely)', ha='center', fontsize=10, style='italic')
    ax.text(2.5, 1.5, 'Low Power, Low Interest\n(Monitor)', ha='center', fontsize=10, style='italic')
    ax.text(7.5, 1.5, 'Low Power, High Interest\n(Keep Informed)', ha='center', fontsize=10, style='italic')
    
    # Sample stakeholders
    stakeholders = [
        {'name': 'Head of Risk', 'power': 8, 'interest': 9, 'color': 'red'},
        {'name': 'CDO', 'power': 9, 'interest': 8, 'color': 'red'},
        {'name': 'Head of IT', 'power': 7, 'interest': 4, 'color': 'orange'},
        {'name': 'Compliance', 'power': 6, 'interest': 7, 'color': 'orange'},
        {'name': 'Data Scientist', 'power': 3, 'interest': 9, 'color': 'blue'},
        {'name': 'Data Analyst', 'power': 2, 'interest': 8, 'color': 'blue'},
        {'name': 'Business User', 'power': 4, 'interest': 6, 'color': 'green'},
        {'name': 'IT Ops', 'power': 4, 'interest': 3, 'color': 'gray'},
        {'name': 'CEO', 'power': 10, 'interest': 3, 'color': 'gray'},
    ]
    
    for s in stakeholders:
        ax.scatter(s['interest'], s['power'], s=200, color=s['color'], alpha=0.7, edgecolor='black')
        ax.text(s['interest'] + 0.3, s['power'] + 0.3, s['name'], fontsize=9)
    
    ax.set_xlabel('Interest')
    ax.set_ylabel('Power')
    ax.set_title('Stakeholder Power/Interest Matrix')
    ax.grid(True, alpha=0.2)
    
    plt.tight_layout()
    plt.savefig('stakeholder_map.png', dpi=300, bbox_inches='tight')
    plt.show()

create_stakeholder_map()
print("Stakeholder map saved as 'stakeholder_map.png'")

# ----------------------------------------------------------------
# PART C: DATA STORYTELLING EXERCISE
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART C: Data Storytelling Exercise")
print("-"*60)

# Simulate some data for storytelling
dates = pd.date_range(start='2023-01-01', periods=12, freq='M')
loan_approvals = np.random.normal(1000, 100, 12).clip(800, 1300).astype(int)
defaults = np.random.normal(30, 5, 12).clip(15, 45).astype(int)
approval_rate = (loan_approvals - defaults) / loan_approvals * 100

df_story = pd.DataFrame({
    'Month': dates,
    'Loan Approvals': loan_approvals,
    'Defaults': defaults,
    'Approval Rate': approval_rate
})

print("Data for Storytelling:")
print(df_story.head())

# Create a storyline
print("\nStorytelling Structure:")

print("""
1. HOOK: "Loan approvals have increased by 15% over the past year, but so have defaults."

2. INSIGHT: "The increase in defaults is driven by a shift towards riskier borrowers."

3. EVIDENCE: 
   - Chart showing approval rates vs default rates.
   - Correlation analysis showing higher defaults among younger borrowers.

4. IMPLICATION: "We need to adjust our risk assessment models to account for this shift."

5. RECOMMENDATION: "Implement a new model that better predicts defaults for younger borrowers."
""")

# Visualise the story
fig, axes = plt.subplots(1, 2, figsize=(14, 5))

# Approval vs Default trend
ax = axes[0]
ax.plot(df_story['Month'], df_story['Loan Approvals'], 'b-', linewidth=2, label='Approvals')
ax.plot(df_story['Month'], df_story['Defaults'], 'r-', linewidth=2, label='Defaults')
ax.set_xlabel('Month')
ax.set_ylabel('Count')
ax.set_title('Loan Approvals vs Defaults')
ax.legend()
ax.grid(True, alpha=0.3)

# Approval Rate
ax = axes[1]
ax.plot(df_story['Month'], df_story['Approval Rate'], 'g-', linewidth=2)
ax.fill_between(df_story['Month'], 0, df_story['Approval Rate'], alpha=0.3, color='green')
ax.set_xlabel('Month')
ax.set_ylabel('Approval Rate (%)')
ax.set_title('Approval Rate Over Time')
ax.axhline(y=95, color='red', linestyle='--', label='Target (95%)')
ax.legend()
ax.grid(True, alpha=0.3)

plt.tight_layout()
plt.savefig('data_story.png', dpi=300, bbox_inches='tight')
plt.show()
print("Data story visualisation saved as 'data_story.png'")

# ----------------------------------------------------------------
# PART D: COMMUNICATION PLAN TEMPLATE
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART D: Communication Plan Template")
print("-"*60)

def generate_communication_plan():
    """Generate a communication plan."""
    plan = pd.DataFrame({
        'Audience': ['Executives', 'Business Owners', 'IT Team', 'Risk/Compliance', 'Data Science Team', 'End Users'],
        'Frequency': ['Monthly', 'Bi-Weekly', 'Weekly', 'Monthly', 'Daily', 'Ad-hoc'],
        'Format': ['Executive Summary', 'Status Report + Demo', 'Technical Review', 'Validation Report', 'Stand-up', 'User Training'],
        'Content': ['Strategic impact, ROI, risks', 'Progress, challenges, decisions', 'Technical design, integration', 'Compliance, validation', 'Task updates, blockers', 'How to use the model'],
        'Owner': ['Project Manager', 'Data Scientist', 'Tech Lead', 'Compliance Lead', 'Team Lead', 'Product Manager']
    })
    return plan

communication_plan = generate_communication_plan()
print("Communication Plan:")
print(communication_plan.to_string(index=False))

# ----------------------------------------------------------------
# PART E: DIFFICULT CONVERSATIONS – SCENARIO PLANNING
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART E: Difficult Conversations – Scenario Planning")
print("-"*60)

scenarios = [
    {
        "Scenario": "Model Performance Declines in Production",
        "Challenge": "The model's AUC dropped from 0.85 to 0.75 over three months.",
        "Stakeholder": "Head of Risk",
        "Approach": "Be transparent, show the data, explain the cause (data drift), present a remediation plan (retrain with new data, adjust thresholds).",
        "Script": "I want to be upfront with you. The model performance has declined due to changes in customer behaviour. We've identified the cause and we're retraining the model. We expect to restore performance within two weeks."
    },
    {
        "Scenario": "Project Delayed Due to Data Issues",
        "Challenge": "Data quality issues have delayed the project by 4 weeks.",
        "Stakeholder": "Project Sponsor",
        "Approach": "Communicate early, explain the reason, present a revised timeline, and propose mitigation steps.",
        "Script": "We've encountered data quality issues that require additional cleaning. We've updated the timeline and added resources to catch up. We expect to deliver by [new date]."
    },
    {
        "Scenario": "Regulatory Rejection of the Model",
        "Challenge": "The model failed the fairness test.",
        "Stakeholder": "Compliance Officer",
        "Approach": "Acknowledge the issue, demonstrate understanding of the requirements, propose corrective actions (re-weighting, debiasing).",
        "Script": "We've reviewed the fairness test results and we understand the issue. We will address it by [specific actions] and resubmit within two weeks."
    }
]

print("Difficult Conversation Scenarios:")
for scenario in scenarios:
    print(f"\nScenario: {scenario['Scenario']}")
    print(f"  Challenge: {scenario['Challenge']}")
    print(f"  Stakeholder: {scenario['Stakeholder']}")
    print(f"  Approach: {scenario['Approach']}")
    print(f"  Script: {scenario['Script']}")

# ----------------------------------------------------------------
# PART F: PRESENTATION TIPS CHECKLIST
# ----------------------------------------------------------------

print("\n" + "-"*60)
print("PART F: Presentation Tips Checklist")
print("-"*60)

tips = [
    "• Start with a strong, clear opening (the 'why').",
    "• Use the Pyramid Principle: start with the conclusion, then arguments, then evidence.",
    "• Limit slides: 1 slide per minute of presentation time.",
    "• Use visuals: charts, diagrams, and images; avoid text-heavy slides.",
    "• Keep language simple and avoid jargon.",
    "• Tell a story: context → conflict → resolution → action.",
    "• Anticipate questions and prepare answers.",
    "• Practice: rehearse with a colleague or in front of a mirror.",
    "• Engage the audience: ask questions, use polls, encourage discussion.",
    "• End with a clear call to action."
]

print("Presentation Tips Checklist:")
for tip in tips:
    print(tip)

# ----------------------------------------------------------------
# PART G: SUMMARY AND RECOMMENDATIONS
# ----------------------------------------------------------------

print("\n" + "="*70)
print("PART G: Summary and Recommendations")
print("="*70)

print("""
Communication and Stakeholder Management – Key Takeaways:

1. Communication is as important as technical skill in data science.
2. Translate technical findings into business language.
3. Use the Pyramid Principle: conclusion → arguments → evidence.
4. Tell compelling data stories: context → conflict → resolution → action.
5. Map stakeholders by power and interest; engage accordingly.
6. Build trust through transparency, reliability, and active listening.
7. Handle difficult conversations with honesty and a solution-oriented approach.
8. Use visualisations effectively: clarity, accuracy, focus, consistency.
9. Tailor communication to the audience: executives, business, IT, risk, peers.
10. Practice, seek feedback, and continuously improve.

Recommendations:
  - Develop a communication plan for each project.
  - Practice data storytelling with real examples.
  - Seek opportunities to present to non-technical audiences.
  - Build relationships with stakeholders proactively.
  - Invest in visualisation skills (course, practice, tools).
  - Join a Toastmasters or public speaking group.
""")

print("="*70)
print("END OF LESSON 5 – MODULE 8")
print("="*70)