This also spurred me to look into my 2018 success and write a blog about that.
First thing out of the way is
I won nothing and it sucks
but the thing I realised about how it sucks here is that the way it sucks is a different suckage from how it used to suck when I entered hackathons. It sucks not because I didn’t win, not because I think my project was better than the other teams’ but mostly because of how hard we worked and how little results came of it. There’s a few ways of looking at this.
-
The idea just wasn’t presetable. I don’t really believe this is the case, as another group did something similar (but just better) and achieved great results.
-
I am too unfamiliar with CNNs and transfer learning. This can be easily solved. I will eventually release a library that does all the heavy lifting for us.
-
Our team composition and my delegation wasn’t good. I really think I lucked into my first team where I didn’t have to worry about whatever aspect that I delegated to. This time I think we should have maybe had one more backend and less FE devs. The AI was also kind of easy to implement maybe someone else would have been faster