Python
import re

def extract_company_info(text):
    info = {}
    
    # Extract CNPJ (Brazilian company ID)
    cnpj_match = re.search(r'CNPJ:\s*(\d{2}\.\d{3}\.\d{3}/\d{4}-\d{2})', text)
    if cnpj_match:
        info['cnpj'] = cnpj_match.group(1)
    
    # Extract emails
    emails = re.findall(r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b', text)
    info['emails'] = emails
    
    # Extract address components
    address_match = re.search(r'(Avenida|Rua|Alameda)[^,]+,[^,]+,[^,]+', text)
    if address_match:
        info['address'] = address_match.group(0)
    
    return info

1. Identify text source - Company footer, legal notice, or contact page
2. Extract structured data:
   • Company name and legal entity type
   • Registration numbers (CNPJ, tax ID, etc.)
   • Physical address with postal code
   • Contact emails by department
   • Phone numbers if present
3. Categorize contacts - Sales, support, legal, etc.
4. Validate format - Check email domains, postal codes, registration numbers
5. Structure output - Organized by contact type and purpose