Skip to content
This repository was archived by the owner on Jan 10, 2026. It is now read-only.

PyPI_DESCRIPTION.md

Latest commit

 

History

History
200 lines (143 loc) · 6.21 KB

PyPI_DESCRIPTION.md

File metadata and controls

200 lines (143 loc) · 6.21 KB

RAIL Score Python SDK

Evaluate and generate responsible AI content with the official Python client for RAIL Score API

PyPI version Python Versions License: MIT

What is RAIL Score?

RAIL Score is a comprehensive framework for evaluating AI systems across 8 critical dimensions:

  • 🛡️ Safety: Content safety and harm prevention
  • 🔒 Privacy: Data protection and privacy preservation
  • ⚖️ Fairness: Bias detection and equitable outcomes
  • 📊 Transparency: Explainability and clarity
  • ✅ Accountability: Responsibility and auditability
  • 🎯 Reliability: Consistency and accuracy
  • 📜 Legal Compliance: Regulatory adherence
  • 👥 User Impact: Positive user experience

This SDK provides easy-to-use Python bindings for the RAIL Score API, enabling you to integrate responsible AI evaluation directly into your applications.

⚡ Quick Install

pip install rail-score

🚀 Quick Start

from rail_score import RailScore

# Initialize client
client = RailScore(api_key="your-rail-api-key")

# Evaluate content
result = client.evaluation.basic("Our AI system ensures user privacy and data security.")

# Access scores
print(f"Overall RAIL Score: {result.rail_score.score}")
print(f"Privacy Score: {result.scores['privacy'].score}")

Get your API key at responsibleailabs.ai

✨ Key Features

🎯 Multiple Evaluation Modes

  • Basic: Quick evaluation across all 8 dimensions
  • Dimension-Specific: Focus on individual aspects (safety, privacy, fairness, etc.)
  • Custom: Choose specific dimensions to evaluate
  • Weighted: Apply custom importance weights
  • Detailed: Get comprehensive breakdown with improvement suggestions
  • Advanced: Ensemble evaluation with higher confidence (0.90+)
  • Batch: Process up to 100 items efficiently
  • RAG: Evaluate RAG responses for hallucinations

🤖 AI Generation

  • Generate context-grounded responses (RAG Chat)
  • Get AI-powered improvement suggestions (Reprompting)
  • Create safe content with minimum quality thresholds (Protected Generation)

✅ Compliance Checks

Built-in support for major compliance frameworks:

  • 🇪🇺 GDPR (EU General Data Protection Regulation)
  • 🇺🇸 CCPA (California Consumer Privacy Act)
  • 🏥 HIPAA (Healthcare data protection)
  • 🤖 EU AI Act (Artificial Intelligence regulation)

🔧 Developer-Friendly

  • ✅ Type-safe: Full typing support with IDE autocomplete
  • 🔄 Auto-retry: Built-in error handling and retries
  • 📊 Detailed metadata: Request IDs, credits, processing times
  • ⚠️ Rich errors: Comprehensive exception handling
  • 📈 Usage tracking: Monitor credits and API usage

📖 Example: Batch Content Moderation

from rail_score import RailScore

client = RailScore(api_key="your-key")

# Evaluate multiple items
items = [
    {"content": "First piece of content"},
    {"content": "Second piece of content"},
    {"content": "Third piece of content"}
]

result = client.evaluation.batch(
    items=items,
    dimensions=["safety", "privacy", "fairness"],
    tier="balanced"
)

# Filter safe content
safe_items = [
    item for i, item in enumerate(items)
    if result.results[i].rail_score.score >= 7.5
]

print(f"Safe content: {len(safe_items)}/{len(items)}")

📖 Example: GDPR Compliance Check

result = client.compliance.gdpr(
    content="We collect user data for personalized recommendations",
    context={"data_type": "personal", "region": "EU"},
    strict_mode=True
)

print(f"Compliance Score: {result.compliance_score}")
print(f"Passed: {result.passed}/{result.requirements_checked}")

for req in result.requirements:
    if req.status == "FAIL":
        print(f"❌ {req.requirement}: {req.issue}")

📖 Example: RAG Hallucination Detection

result = client.evaluation.rag_evaluate(
    query="What is the capital of France?",
    response="The capital of France is Paris, located in the Île-de-France region.",
    context_chunks=[
        {"content": "Paris is the capital and largest city of France."},
        {"content": "The Île-de-France region surrounds Paris."}
    ]
)

metrics = result['result']['rag_metrics']
print(f"Hallucination Score: {metrics['hallucination_score']}")  # Lower is better
print(f"Grounding Score: {result['result']['grounding_score']}")  # Higher is better

🎓 Use Cases

  • Content Moderation: Evaluate user-generated content for safety and compliance
  • AI Model Evaluation: Assess LLM outputs across responsible AI dimensions
  • Compliance Automation: Automated GDPR, HIPAA, CCPA checks
  • RAG Quality Assurance: Detect hallucinations in retrieval-augmented generation
  • Policy Enforcement: Ensure AI-generated content meets organizational standards
  • Research & Development: Benchmark responsible AI metrics

📚 Documentation & Resources

🔧 Requirements

  • Python 3.8 or higher
  • requests >= 2.28.0

📄 License

MIT License - see LICENSE for details.

🤝 Support

Made with ❤️ by Responsible AI Labs