I have been on enterprise incident calls with a hundred people on the line, everyone taking turns asking the same question: is it your problem? Is it your problem? That is not a root-cause process. That is a dashboard failure with an audience.
The logging and the dashboards exist for one long-term reason: when something breaks, get to the root cause as fast as possible. A single pane of glass should tell you every system is green — and when one is not, you drill into it, and keep drilling, until you are looking at the actual issue. If the pane cannot take you from red to root cause, it is a status page, not an instrument.
And the pane is no longer just for humans. In a world of agentic systems, the more you track, the easier it is to create triggers that spawn an AI agent the moment the platform senses an issue: go look, understand it, potentially make the fix, submit the PR. Resolved in minutes, before anyone even realizes there was a problem.
That is the goal I build toward. A hundred people on a call is what it looks like when a system cannot explain itself. Track everything, and it can.
More writing
Why the most autonomous system I run got there through telemetry, lineage, and gates, not smarter models
Newer · Aug 2026
Why the résumé I read is the systems catalog, not the badge list
Older · Aug 2026
Start
I’ll tell you in about a day whether I’m the right person. The first conversation is fit, not a free architecture review.