Agaryan

Thanks for the very complete answer!

In my previous company we followed a pretty similar process. With the five whys and next step actions in our post-mortems. We built a in-house app organise everything during the crisis and generated the post-mortem on confluence afterwards. But it was a lot of engineering and it's not fit for a small start-up bootstrapping its product.

Okay, thanks for your answer! So you're using pagerduty and defined your own incident process with priorities and roles. Interesting. When a crisis arises during the day, the whole team is meeting in a war room and focus on fixing it or only the on-duty people work on it?