Causal inference and probabilistic machine learning at scale
I am a research scientist at the Basis Research Institute
(New York/Cambridge, MA), a nonprofit AI research-and-engineering lab building a
general-purpose reasoning engine. I work on causal inference and probabilistic machine
learning, and on making them run at scale, from formal method to production system. I am
a core contributor to ChiRho, Basis’s
open-source causal probabilistic programming language (built on Pyro/PyTorch). Among
other things, I lead the work on its
Explainable module, the
probabilistic causal impact attribution machinery.
My research builds causal Bayesian models with deep machine-learned components for attribution and counterfactual reasoning under uncertainty in complex, partly observed dynamical systems, with a recent emphasis on spatio-temporal problems. One recent paper gives a counterfactual semantics for hybrid dynamical systems (NeurIPS 2025).
I hold a PhD in Logic and Philosophy of Mathematics from the University of Calgary, and have held positions at the Research Foundation – Flanders, Trinity College Dublin, the University of Bristol, and the University of Gdańsk.
Currently working on
- Spatio-temporal causal modeling. Causal Bayesian models with deep machine-learned components for spatio-temporal valuation problems, dynamical models of policy interventions, and causal models of group foraging behavior in animals. The shared question is how to make causal explanation tractable when the system is high-dimensional, dynamical, and only partly observed.
- PCI (Probabilistic Causal Impact). A causal attribution method that decomposes explanations into separate necessity and sufficiency components instead of collapsing them. It bridges Halpern–Pearl actual causality with scalable attribution machinery (SHAP, LIME, gradient-based methods) by re-casting explanation as estimation on an expanded probabilistic causal model.
- LLM interpretability at scale. Combining probabilistic causal impact tooling with LLM architecture to scale up interpretability work, moving from one-off probing toward systematic, causal accounts of what drives model behavior.
- Common Task Framework for scientific ML. Collaborating with the Common Task Framework team on shared benchmarks for scientific machine-learning methods. Curated datasets and task-specific metrics (forecasting, state reconstruction, generalization, and control on canonical dynamical systems) let us compare methods head-to-head.
- Model and data skepticism for epidemiology. Developing algorithmic methods for automated skepticism about epidemiological models and the data they rest on, stress-testing both to find where conclusions come from fragile assumptions or data-quality artefacts rather than signal.
Open source
- ChiRho ★ 269.
Basis’s open-source causal probabilistic programming language, built on Pyro/PyTorch.
I am a core contributor and lead its
Explainablemodule. - collab-creatures. Lead developer; the Bayesian and causal modeling toolkit behind our group-foraging research (Scientific Reports, 2024).
Most of my engineering work happens in private repositories. The chart below summarizes my GitHub activity by year: pull requests authored and merged (top), reviews and issues (bottom), across public and private repositories.
GitHub activity per year: pull requests authored and merged (top); reviews and issues (bottom). Public and private repositories combined; 2026 is year-to-date.