This is going to be the first post in a series(I hope) about the development of this new app I am working on. The project is currently operating under the name "Yaquiza" and perhaps this will be the name of the app when it's launched but I am not quite sure. If I do decide on this name though, I am quite prepared since I already own the yaquiza.com domain 😬.

Why am I starting this series?

Firstly, if you haven't already noticed, this "blog" is quite the wasteland. It could use some words and the only way to do that is by writing.

Secondly, I have anyways wanted to get some tech-related writing done. Can't think of anything better than to talk about the project that I hope to turn into a money-making venture some day.

Thirdly and finally, it's nice to take stock of things on a monthly basis. I had actually started working on a few skeleton screens a long time ago(I am too embarrassed to mention the precise time) but it's only this year(and this month) that I have gotten around to building some momentum. So I thought that maybe if I recall the progress I have made every month and write about it, I will be able to see the impact I have been able to make by just working 1-2 hours a day every day. And hopefully, this will keep me going forward.

What is the app about?

It's a quiz app.

While I have always been a big trivia/quizzing buff, my interest in the area was reinvigorated in 2021 when I became aware of JetPunk.com. I love the simplicity of the website and the catalog of quizzes it provides(especially the Maps and Flag quizzes).

I was surprised to see that JetPunk had been around for so long and I had only just become aware of it. I was also surprised to see that they did not have a mobile app. I understand the reasoning behind it and to be fair, the mobile-website works pretty well. But I do know that if they did have a mobile app, I would have installed it and I would be using JetPunk more than I do right now.

There are however alternatives to JetPunk for mobile(ex: Sporcle) but most of them feel very clunky and slow. And a lot of these don't work offline.

Additionally, none of these quiz apps have managed to capture the magic like QuizUp did back when it was operational.

Based on the history of this genre of apps, my understanding is that while there is a sizeable audience for such apps, there isn't bucketloads of money to be made. My idea is that by picking the best features from the various apps, I can create an app that will hopefully be the ultimate quiz app.

It will be fast, it will have a great catalog of quizzes, large parts of it will run offline and because of that, it will be cheap to run. The hope and expectation is that with all these ingredients coming together, the users will also come and then one day money too.

Only one way to find this out though. By building the thing.

What's the tech stack?

I am starting with the Android app first. And for that I am using:

For the iOS app, my intention is to use Kotlin Multiplatform + Compose Multiplatform for iOS to enable as much codesharing as possible between the 2 platforms.

In retrospect, React Native would have been a solid choice for the mobile apps too, given the kind of app I am building and the fact that I want to share as much code as possible between the 2 platforms. But oh well.

For the server side, I am starting with:

Kotlin
Ktor
Kotlin Coroutines
SQLDelight
PostgreSQL

While I have done some backend work in previous jobs, I have never actually built anything from scratch and actually launched it to loads of users. I haven't quite made up my mind on where/how to host the server. Do I embrace the cloud? Or do I go the VPS route?. Let's see where this road takes me.

Additionally, I will also build a desktop app for admin users. This will use Compose Multiplatform for Desktop for the UI and because it will also be a JVM app, a lot of the tooling of the Android app would carryover here. My hope is that I could eventually have a single codebase to build both the admin as well as public facing apps.

All in all, nothing too controversial here I think. Just sticking to the stuff that I know I am extremely comfortable with.

What all had been built before February 2025?

Truly, not much.

The Android app would launch at the home screen showing a list of MCQ(Multiple Choice Question) quizzes where the data was hardcoded. You could open a quiz in the quiz taking screen which was also functional. But only MCQ Quizzes are supported for the moment.

There was also the registration and login screen which had the UI implemented however it had not been tested or hooked with the other 2 screens.

And, that's it. Like I said, not much.

What was the progress in February 2025?

On the app side, the registration and login flow was hooked with the other 2 screens. All of the data is saved locally currently but besides that it works as expected.

On the server side, the first code was written this month and I have put together the DB design, implemented auth and just finished implementation of the API to get quizzes. The DB is currently an in-memory SQLite DB which is good enough to get testing started. And for testing, I have used Postman to ensure everything is working as expected.

While it may not look like a lot, I am quite pleased with the progress. The setup of the server side code is decent and I feel like I am finally past the stage where I have to stop and think about every line of code I am adding. Going forward, my pace should increase and hopefully I will be more productive.

How did I use LLMs?

I am still at a stage where I am still figuring out what are the best use cases for LLMs in my daily life. I don't know where one model does better than the other nor am I interested in answering that question at this point. So using ChatGPT, I was pleased to be able get the following done:

Generating regular expressions for user input validation of emails, usernames and passwords.
Generating SQL queries.
Generating dummy data for testing.

These 3 items describe use cases where I am already capable of performing the tasks but it would just take longer and the output I would generate would likely not be any better than what ChatGPT is capable of generating. Additionally, I would be able to critique the output generated and make corrections where needed. And the need for corrections did arise, for example in a regular expression it generated, to check if a password matched the expected requirements. I tried to push ChatGPT to make the correction but it continued to spit out an incorrect regular expression. Not sure if I need to up my "prompt engineering" skills.

I will also use ChatGPT in writing tests(whenever I get around to it 🙈) but ultimately even in that case, I will be the final authority to decide what goes in and what stays out.

And I think that's the best kind of relationship to have with it for the moment.

Goals for March 2025

I have 2 APIs that I still need to implement for the flows already implemented on the user mobile app. These should definitely be ready for use hopefully in a week or so.

Once that is done, I will move to integrating all the APIs on the user mobile app. I think integrating all the APIs by the end of March would be a reasonable target. Essentially, I should have an app that is able to run on a local server end to end.

Once that is done, I will move to building the admin desktop app which will have flows for user verification and quiz creation. Once these 2 flows have been implemented and the end-to-end experience works correctly across the desktop and mobile apps, I will consider v0.0.1 complete.

That's it for now, see you in a month ✌️

Addendum

To read about the monthly update for March 2025, click here.