The books
Assessments we take on
Each assessment is a piece of human work with a start, a document, and an end. None of them replace your monitoring stack. They read what that stack, and the people around it, already produce.
Choose the sitting that matches the problem you can already name. If the problem is still vague — “the pager is loud” is a start — begin with the Application Health Review. The other three sit beside it when a narrower question is already on the table.
Alerting Hygiene Sitting
A focused two-day sitting that walks the live alert list with the people who still receive the pages, then marks which conditions belong on the night phone.
Application Health Review
A twelve-to-fifteen-day sitting over one production application: interviews, observation of alerting, ninety days of incident records, a written health brief, and a readout.
Incident Record Reading
A close reading of a bounded set of incident records to recover what the health signals did, and did not, show before each event.
Retained Observation
A quarterly return to the same application: a short observation window, a comparison against the last brief, and a note on what the health signals have done since.