Deal with really big log files, mostly.

If you work at a hyperscaler, service log volume borders on the insane, and while there is a whole pile of tooling around logs, often there's no real substitute for pulling a couple of terabytes locally and going to town on them.