Quick start notebook for testing sensitive-information guardrail

[yoshitafiddler]

Comprehensive Quickstart Notebook Created!

I've created a well-structured Jupyter notebook (sensitive_information_guardrail_quickstart.ipynb) that demonstrates how to use the sensitive information guardrail with all the new entity configuration features.

What's Included:

🔧 Core Features:

• ✅ Complete API setup with guardrail credentials
• ✅ Configurable entity detection function with PII/PHI/Custom support
• ✅ Pretty-print results with entity grouping and confidence scores
• ✅ Error handling and robust API calls

Examples Covered:

• PII Detection (Default) - Personal, financial, and digital identifiers
• PHI Detection - Healthcare-specific entities (medications, conditions, insurance)
• Combined PII + PHI - Comprehensive healthcare compliance scanning
• Custom Entities - Enterprise-specific sensitive data detection
• Batch Processing - Efficient processing of multiple texts with performance metrics

Configuration Templates:

• Financial screening (PII only)
• Healthcare compliance (PII + PHI)
• Enterprise security (PII + Custom)
• Comprehensive audit (All entity types)

🚀 Key Benefits:

• ✅ Ready to Run: Just execute the cells to see live examples
• ✅ Production Ready: Includes error handling, retries, and batch processing
• ✅ Flexible: Easy to adapt for your specific entity categories
• ✅ Educational: Clear explanations and best practices
• ✅ Comprehensive: Covers all new entity configuration features

🎯 Usage Examples:
The notebook is structured for both learning and production use.

# Basic PII detection
response, latency = get_sensitive_information_response(text)

# Healthcare compliance (PII + PHI)
response, latency = get_sensitive_information_response(
    text, entity_categories=['PII', 'PHI']
)

# Custom enterprise entities
response, latency = get_sensitive_information_response(
    text, 
    entity_categories='Custom Entities',
    custom_entities=['employee id', 'api key', 'project code']
)

# Batch processing for datasets
results_df = process_dataset_batch(
    text_list, entity_categories=['PII', 'PHI'], max_records=100
)