AI-Powered Data Analytics with Anthropic Claude

# Step 1: Install Anthropic SDK
# In a Google Colab or Jupyter notebook cell, run the following:
!pip install anthropic

# Step 2: Import and configure
import anthropic

# Initialize the Anthropic client
# IMPORTANT: Replace 'YOUR_API_KEY' with your actual Anthropic API key
client = anthropic.Anthropic(
    api_key="YOUR_API_KEY"
)

print("Claude SDK initialized successfully!")

# Step 3: Create sample dataset
import csv
from io import StringIO

# Sample sales data (12 months, 3 regions, 4 products)
data = """Date,Region,Product,Units_Sold,Revenue,Cost
2025-01,North,Laptop,145,217500,130500
2025-01,North,Phone,289,144500,86700
2025-01,North,Tablet,178,53400,35560
2025-01,North,Headphones,412,41200,20600
2025-02,North,Laptop,158,237000,142200
2025-02,North,Phone,301,150500,90300
2025-02,North,Tablet,192,57600,38400
2025-02,North,Headphones,445,44500,22250
2025-03,North,Laptop,172,258000,154800
2025-03,North,Phone,334,167000,100200
2025-03,North,Tablet,203,60900,40590
2025-03,North,Headphones,478,47800,23900
2025-01,South,Laptop,98,147000,88200
2025-01,South,Phone,187,93500,56100
2025-01,South,Tablet,134,40200,26780
2025-01,South,Headphones,298,29800,14900
2025-02,South,Laptop,112,168000,100800
2025-02,South,Phone,203,101500,60900
2025-02,South,Tablet,145,43500,28980
2025-02,South,Headphones,321,32100,16050
2025-03,South,Laptop,127,190500,114300
2025-03,South,Phone,224,112000,67200
2025-03,South,Tablet,156,46800,31200
2025-03,South,Headphones,345,34500,17250
2025-01,West,Laptop,187,280500,168300
2025-01,West,Phone,356,178000,106800
2025-01,West,Tablet,223,66900,44580
2025-01,West,Headphones,534,53400,26700
2025-02,West,Laptop,201,301500,180900
2025-02,West,Phone,389,194500,116700
2025-02,West,Tablet,241,72300,48210
2025-02,West,Headphones,578,57800,28900
2025-03,West,Laptop,219,328500,197100
2025-03,West,Phone,423,211500,126900
2025-03,West,Tablet,267,80100,53400
2025-03,West,Headphones,612,61200,30600"""

# Parse the CSV to verify it loads correctly
csv_reader = csv.DictReader(StringIO(data))
rows = list(csv_reader)

# Display first 5 rows
print("Dataset Preview (First 5 Rows):")
print(f"{'Date':<12} {'Region':<10} {'Product':<12} {'Units':<8} {'Revenue':<10} {'Cost':<10}")
print("=" * 70)
for row in rows[:5]:
    print(f"{row['Date']:<12} {row['Region']:<10} {row['Product']:<12} "
          f"{row['Units_Sold']:<8} ${int(row['Revenue']):>8,} ${int(row['Cost']):>8,}")

# Step 4: Get AI analysis summary
# Note: Ensure you ran the previous cells to define 'client' and 'data'

summary_prompt = f"""
I have a sales dataset with the following data:

{data}

Please provide a comprehensive analysis including:
1. Total revenue and profit across all regions
2. Best performing region and product category
3. Month-over-month growth trends
4. Any notable patterns or anomalies

Format your response in clear, business-friendly language.
"""

message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=2048,
    messages=[
        {"role": "user", "content": summary_prompt}
    ]
)

analysis = message.content[0].text
print("=== CLAUDE'S EXECUTIVE SUMMARY ===\n")
print(analysis)

# Step 5: Deep pattern analysis
pattern_prompt = f"""
Based on this sales data:

{data}

Please identify:
1. The top 3 product-region combinations by revenue
2. Which region has the most growth potential (lowest current performance)
3. Are there any concerning trends in specific product categories?
4. Which product should we focus on to maximize profit?

Be specific with numbers and explain your reasoning.
"""

message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1536,
    messages=[
        {"role": "user", "content": pattern_prompt}
    ]
)

patterns = message.content[0].text
print("=== DEEP PATTERN ANALYSIS ===\n")
print(patterns)

# Step 6: Strategic recommendations
recommendation_prompt = f"""
As a business analyst reviewing this Q1 data:

{data}

Provide 5 specific, actionable recommendations for Q2 2026 that would:
- Increase total revenue by at least 15%
- Improve profit margins
- Address regional performance gaps

For each recommendation, explain the expected impact and why it would work.
"""

message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=2048,
    messages=[
        {"role": "user", "content": recommendation_prompt}
    ]
)

recommendations = message.content[0].text
print("=== STRATEGIC RECOMMENDATIONS ===\n")
print(recommendations)