Post mortem room

A post mortem is the look back a team takes once a project is over, or after something has gone wrong: what happened, why it happened and what to change next time. The name is borrowed from medicine, but nothing has to have died. Good projects get one too.

Its job is to get every issue on the table, including the ones nobody would raise out loud. Done the usual way it is slow. The write-up waits days or weeks for one person to find the time to collect everyone's part and pull it together. The post mortem room does all of it in a one-hour meeting.

The team sits in one room, each person on their own laptop or phone, and joins with a six-character code. Everyone writes at the same time instead of taking turns to speak, so the first voice does not steer the rest and nobody waits for anyone to finish their part. The board walks the room through what went well, what went wrong, why it happened, the lessons and the actions, with an owner and a date on each action.

AI is used only where it helps. It suggests lessons and actions, turns an uploaded schedule into a draft timeline, and writes the report, tracing each problem back to its cause. The room votes on its suggestions and the facilitator decides what is kept.

When the board is full, one click writes the report and it streams onto every screen. The post mortem is finished before anyone leaves the room.

• It is blameless by rule. The AI may not invent anything or name a person as a cause. Names appear only in the list of who attended and beside the actions they own.
• Votes sort the board and order the report, which never prints the counts.
• Notes link back to what they explain, so each root cause is written up under the problem behind it.
• The facilitator can bring every screen to the same step and run one timer for the room.
• Join codes leave out I, L, O, 0 and 1, so they read cleanly off a projector.

The session in these pictures is invented.