Not necessarily. There might be some things that are already set in stone. E.g. the companies build and version control system might be fixed, you might be deem

Not necessarily. There might be some things that are already set in stone. E.g. the companies build and version control system might be fixed, you might be deemed to use protobufs, etc. But there can still be one area to innovate on! It's usually not productive trying to reinvent everything anyway. Pick one thing, and excel at that. There are enough of those challenges also at bigger companies. Maybe not at every team, but there is likely enough internal flexibility that one can find a suitable team. And really, inventing the 100th iteration of a custom CI or RPC system at a small company isn't that exciting either.