Software

I develop software that supports my research in causal inference, sequential decision-making, and machine learning, ranging from reusable packages to research code released alongside specific papers.

Open-source libraries:

med_bench is a Python package for causal mediation analysis that implements the most common estimators within a unified framework, while allowing flexible choices for nuisance parameter models. It accompanies the paper Causal mediation analysis with one or multiple mediators: a comparative study.

Open-source research software:

• adaptive-KTE (2026)
• counterfactual-policy-mean-embedding (2025)
• double-debiased-machine-learning-mediation-continuous-treatments (2025)
• sequential-conterfactual-risk-minimization (2023)
• Efficient-Kernel-UCB (2022)
• Nested-Exponential-Weights (2022)
• optimization-continuous-action-crm (2020)
• Adversarially-Learned-Anomaly-Detection (2018)
• Efficient-GAN-Anomaly-Detection (2018)