XFIN-XAI: Privacy-Preserving Explainable AI for Financial Services

XFIN-XAI is an open-source Python library designed for privacy-preserving explainable AI (XAI) in financial services and banking systems. It enables banks and financial institutions to generate transparent explanations for black-box models without exposing proprietary internals.

The library focuses on credit risk explanations, adverse action notices, and counterfactual recommendations, ensuring compliance with regulations like GDPR and ECOA.

Note: This library is built for educational and research purposes, allowing users to explore XAI in finance while maintaining data privacy.

🚀 Features

• 🔒 Privacy-Preserving Explanations: Integrates SHAP and LIME for local explanations using only model predictions
• 💳 Credit Risk Module: Generates feature importances, adverse action notices, and actionable recommendations
• 📋 Compliance Engine: Produces regulatory-compliant reports and audit trails
• 🤖 LLM Integration: Uses Gemini (or similar) for natural language explanations and recommendations
• 🔧 Modular Design: Easily extensible for other domains (e.g., ESG, stress testing)
• ⚡ Efficient and Scalable: Runs on commodity hardware with low resource usage

📦 Installation

Quick Installation

pip install xfin-xai

Launch the Web Interface

After installation, launch the interactive web interface:

xfin credit

This will open the Streamlit web application where you can upload your model and dataset files.

Command Line Options

# Show help
xfin credit --help

# Launch on custom port
xfin credit --port 8502

# Launch on all interfaces
xfin credit --host 0.0.0.0

Development Installation

For development installation:

git clone https://github.com/dhruvparmar10/XFIN.git
cd XFIN
pip install -e .

🔧 Requirements

• Python: 3.9+
• Core Dependencies: pandas, numpy, shap, lime, joblib, matplotlib, streamlit, scikit-learn
• Optional: OpenRouter API key for LLM-powered explanations

See requirements.txt for the complete list.

🚀 Quick Start

Here's a basic example to get started with credit risk explanations:

import pandas as pd
import joblib
from XFIN import CreditRiskModule

# Load your black-box model
model = joblib.load('path/to/your/model.pkl')

# Define a wrapper for your model (only expose predict/predict_proba)
class BankModel:
    def predict(self, X):
        return model.predict(X)

    def predict_proba(self, X):
        return model.predict_proba(X)

# Sample input data (replace with your features)
sample_data = pd.DataFrame({
    'Annual_income': [50000],
    'Employed_days': [1825],
    'Credit_score': [650],
    # Add other features as per your dataset
})

# Initialize the explainer with API key (optional)
explainer = CreditRiskModule(
    BankModel(),
    domain="credit_risk",
    api_key="your-openrouter-api-key"  # Optional for LLM explanations
)

# Generate explanation
explanation = explainer.explain_prediction(sample_data)

# Generate recommendations
recommendations = explainer.generate_recommendations(sample_data)

# Generate compliance notice
compliance = explainer.generate_adverse_action_notice(explanation)

print("Prediction Explanation:", explanation)
print("Recommendations:", recommendations)
print("Adverse Action Notice:", compliance)

📖 Documentation

Full documentation is available at xfin-xai.readthedocs.io.

• 📚 API Reference
• 🎓 Tutorials
• 🗺️ Roadmap

🤝 Contributing

We welcome contributions! Please see our CONTRIBUTING.md for guidelines.

How to Contribute

1. Fork the repository
2. Create a feature branch (git checkout -b feature/YourFeature)
3. Commit your changes (git commit -m 'Add YourFeature')
4. Push to the branch (git push origin feature/YourFeature)
5. Open a Pull Request

For bugs or feature requests, please open an issue on GitHub.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• SHAP and LIME teams for building excellent explainability tools
• Open-source community for tools like setuptools and ReadTheDocs
• Financial AI research community for guidance on regulatory compliance

📞 Contact

For questions or support:

• Email: dhruv.jparmar0@gmail.com
• Issues: GitHub Issues