Root-cause analysis: logs, metrics and deploys in a single picture
On an incident signal the scenario collects logs, metrics, deploys and flags, clusters the errors, lines the events up in time and posts the analysis to chat.
Four sources collected at once instead of one by one, by hand
A thousand log lines collapse into a handful of failure types
Events lined up in time — you can see what came first
The analysis lands in chat instead of staying in one person's head