Index · 6 posts
Writing
Notes on test infrastructure, structured LLM extraction, and reaching for the legible solution first. Drafts are marked; full write-ups are in progress.
-
What the reviewer-recommendation engine taught me about reaching for the legible solution first.
-
Using models to turn messy test output into clean, queryable data — the boring, useful application.
-
Why you have five note apps: a three-axis model — object type × verb × register — of the whole space personal-information tools occupy, and the orchestration layer that belongs between them.
-
The most reliable place to put a model is between two systems that already know what they want.
-
Notes on a compliance-data pipeline — scraping, normalizing, and scoring 10,000+ establishments.
-
Why formal ontology — not better prompts — might be the fix for AI agents that write plausible but wrong code.