Interpretability-first ML pipeline wrapper for regulated industries.
Train any scikit-learn / XGBoost model and get SHAP explanations, counterfactuals, an audit log, and a compliance report (SR 11-7, EU AI Act Art.13, FINRA) -- automatically.
from interpretflow import InterpretableModel
model = InterpretableModel(base_model="xgboost")
model.fit(X_train, y_train)
report = model.compliance_report(standard="SR_11_7")Status: Phase 1 in active development.
pip install interpretflow
pip install "interpretflow[full]" # + counterfactuals, PDF reportsSee CONTRIBUTING.md. PRs welcome!
Apache 2.0 -- see LICENSE.