I built a lot of weird 'inventions' in university, such as an airsoft sentry gun to keep squirrels away from the garden. I was also in a student club.
I don't think any project specifically made potential employers say 'wow, we definitely have to hire this guy', but I think having actual experience doing stuff and demonstrating an interest for the field goes a long way towards landing you those first couple of jobs.
I noticed that when you propose a new idea, often there is no shortage of resident "experts" giving you hand-wavy reasons as to why it can't work, won't work, etc. I kind of feel like you DO need a PhD just to have enough clout for these kinds of people to trust you.
Here's one I find doesn't get enough credit: Reading comprehension.
That can be interpreted as reading someone else's code, but even more importantly: Just being able to understand plain English (or whatever human language).
I've seen so much code that used an API incorrectly, or misunderstood (or simply ignored) documentation. Being able to RTFM, imo, is essential. See also: Understanding what your boss wants and delivering it.
I don't think any project specifically made potential employers say 'wow, we definitely have to hire this guy', but I think having actual experience doing stuff and demonstrating an interest for the field goes a long way towards landing you those first couple of jobs.