Rinum

Summary

A primitive chatGPT that uses user-given answer to answer questions. An assignment in my Algorithm Strategy subject.
P.S. The site sadly no longer works because the database it uses have been deleted.

Developer Diaries

I didn't make this alone, instead I was part of a team of three. Although I am the one who worked on the website side of things from the UI/UX to the FE and the BE.

Library-Accelerated Development

Making this only took me about 3 days, and they're not a whole day since I was on lebaran holiday at the time. It only took me that long because of how the library I used really made the development fast. DaisyUI and Tailwind enabled me to make something as decent-looking as this with little effort. Next Auth made authentication as simple as modifying a file (though I had to spend a bit wondering why Google authentication wasn't instantly working). Prisma prevented any human error of writing the wrong query to the database and made me faster in writing those queries. React Query (or Tanstack Query as it is called now) prevented any headache over asynchronous state management.

The (Some) Corners Cut

I'd be lying if I say the secret to making this in such a short time is simply by using library. This site is approximately 90% of what I think it could be. Not doing the extra 10% is one of the reason the development can be done quickly. The 10% are the similarity to the real ChatGPT (if you notice, certain things such as session creation works a bit differently, and how the response are returned is quite clearly different) and creature comforts such as optimistic updates. I did this because I had a lot of other things to do at the time and wanted this to get out of the way as quickly as possible.