Mostly I count things.
Sometimes I go on to say which number is larger. Applied causal inference. Practicing probabilistic alchemy. Occasionally the numbers are wrong.
I build experimentation programs, data products, and agentic systems. Causal inference, Bayesian analysis, and the infrastructure between data and decisions.
It's never been cheaper to ship new features—but knowing what to build is as hard as ever. The bottleneck has moved entirely to measurement: what worked, for whom, and what to try next.
I make systems that answer that question—for both humans and the agentic approximations operating alongside them. Structured causal frameworks that measure real treatment effects, detect who responds differently, and feed that learning back into the next decision—consistently and automatically.
AI Agents Are an Infinite Queue of Fresh CS Grads with Eidetic Recall of Stack Overflow That You Shoot in the Head After They Complete Their First Ticket
Why LLM-generated code often creates more problems than it solves, and what compile-time controls can do about it.
My Current Claude Franken-Workflow
A walkthrough of the spec-driven, orchestrator-subagent system I've built for working with Claude Code on a real codebase.
Some Knowledge Only Comes From Making the Thing
The difference between producing software and understanding it, and why building teaches you something you can't get any other way.