~/writing
Writing
Essays and notes on causal inference, machine learning, AI evaluation, and building a technical company — written for both technical and non-technical readers.
Founder Notes4 min
Building Censiq: Notes From the First Version
Founder notes on why evaluation infrastructure for AI agents needed to exist as its own product, not a feature bolted onto something else.
AI Evaluation5 min
What 'Evaluating an AI Agent' Actually Means
Most AI-agent evaluation stops at a benchmark score. Role-specific evaluation asks a harder, more useful question: would this agent succeed at the actual job.
Causal Inference6 min
Prediction and Causation Are Different Questions
A model that predicts well is not the same as a model that tells you what to change. Why that distinction matters for anyone making decisions from data.