I suspect the next big thing will be to incorporate machine learning into the software development life cycle. - Issue comments that are no longer relevant,

I suspect the next big thing will be to incorporate machine learning into the software development life cycle. - Issue comments that are no longer relevant, should be collapsed by default. - You should be able to get a "too long; did not read" for issue comments. - Searches should be context aware. Matches should take into consideration the user, time, branch (is this a release branch, developer branch, shared development branch, etc.) and so on. - Build logs should include a "too long; did not read" - Builds should be context aware to prevent wasting of resources. I guess what ever can save time, money and speed development would be next generation features. I'm not sure what GitLab's sales engagement looks like, but if you are not, you should be asking to speak with a company's internal, tools dev team, if you can. Large companies spend an insane amount of money and time, creating/maintaining bespoked dev tools for internal use. Companies like Google, Facebook, Twitter, etc. spend millions in labour cost annually, to build and maintain their internal solutions. GitLab's goal, should be to see if they can create comparable solutions, that can be managed by small startups and mid size companies. Having worked in tools dev for Enterprise, I knew being able to search and analzye anything, is a time saver, but I also knew, it would be cost prohibitive for many smaller companies to provide such a solution internally. This was why I set out to develop a solution that could be maintained by companies of all sizes and this should be the thought process for GitLab, in my opinion. If GitLab wants to work on solving the harder problems, it should understand what companies are spending internal resources on creating/maintaining and see if they can't make it available for the masses. What you learn could lead to more intelligent code review solutions among other things. In a lot of ways, you are trying to capture somebodies skill set/knowledge, which isn't all that trivial.