I work on causal AI — the part of machine learning that cares less about prediction and more about consequence. Most of my time goes into causal inference, uplift modeling, and turning research-grade ideas into systems that hold up in production.
Currently
Building tooling around causal discovery and incremental treatment effects; writing the occasional note; giving the occasional talk.
Interests
- Identification and the quiet assumptions behind every estimate
- Uplift / heterogeneous treatment effects
- Graphs, structure, and the geometry of cause
- Minimal, fast, content-first software
Elsewhere
Find the outbound signals on the links page — GitHub, LinkedIn, and the rest.
Correlation is loud. Causation is quiet. This page is mostly the latter.