🔬 Stylometry Engine

🧪 Starter content — edit src/content/ai-projects/stylometry-engine.md.

A research-driven pipeline that turns a corpus of an author’s published work into a quantified voice profile — function-word distributions, sentence-rhythm statistics, punctuation habits — and then uses that profile to steer LLM drafts toward the author’s real style instead of generic model defaults.

Status

Work in progress. The measurement layer is solid; the feedback loop into drafting is the active research question.

How it connects to the papers

This is the applied side of my corpus and authorship research — work like Gender Effects in Student Technical Writing makes the methodological case; this project is the running code.