Machine Learning for issues
I get pinged on so many issues. There are so many things I should be reading. There are too many issues for me to go through alone. We should use machine learning to figure out if an issue is important, what the gist of the issue is, and make a legit todo for people. This is a huge problem for leads. We get notified on so many issues we just ignore stuff after a while.
Just like Gmail knows if an email is important. Everything is important at GitLab. But some things are more important than others. And it might know who to ping as well. So if you are writing about issue boards, it would be wise to involve @iamphill in the conversation even though he is now on the AC team. And mentions of frontend should probably mention Tim or I. A suggested list of people to mention and then a customized list of todos for people showing them what they should act on first.
A customized todo list based on machine learning would be amazing.