If you’ve ever stared at an ugly Excel spreadsheet at 2:00 AM trying to figure out how to save for your new home or even a LEGO set, you already know why traditional budgeting apps make people want to throw their phones out a window. Half of the apps feel like boring homework, and the other half just blast ads in your face the second you open them. I have noticed these things happen way too many times, and I thought to solve it with …drum roll… PocketVault. So, I cracked my knuckles and switched on my laptop. But every developer knows the feeling – you get a flash of inspiration, you see the app perfectly in your head, and then… you just stare at a blank IDE. That was me with Pocket Vault. I wanted a seamless way to track budgets, but bridging the gap between an idea and a shipped product was my first real challenge. Just getting started was half the battle. So, I downloaded Xcode and began coding away endlessly….. WAIT, I’m joking; I did not do that. I actually made a very quick prototype of PocketVault and posted that on YouTube, Facebook, Reddit, and Discord to see the response and whether the audience liked it or not.
Feedback I got to improve my prototype
Views I got from Reddit, Facebook and Youtube combined for my first prototype of Pocket Vault
I couldn’t believe what I was seeing with my own eyes; I had over 2k + counting views combined. That was a lot for me since this was my first time posting on any of these platforms (I know what you guys are thinking – “But Aarav, that’s nothing. My basic videos get over 100k views. I’ll just say that social media has ruined the perspective of numbers.), and I even got a few comments on how to improve my prototype.
So that lit a fire in me. I worked endlessly to redesign the UI (adding my own colour contrast in the app, making dark and light mode, etc.) for a week to improve my prototype and bring it one step closer to the actual product. Then I rolled out the video showing off version 1 of my actual app and how I improved the ui of prototype, like everyone was asking. Then naturally I started looking at all the different platforms to post my version 1 on. Not much choice except the App Store, or at least that’s what I thought at first. I looked at the price to post it on the App Store and yea… I ain’t gonna pay $99 USD / year as a student for an Apple developer account for my very first app. So I was very disappointed. One day, while I was brushing, I thought that “Apple can jump off the cliff (metaphorically, obviously), and let’s put my app on the PlayStore and Samsung Galaxy Store.” So, I hopped on Google, hoping that posting on either of those was less than that bonker of a price. As soon as I hit the search button, I genuinely got blessed because right in front of me, Google was saying that posting on Galaxy Store was absolutely free and PlayStore only required $22 USD ONCE. I was going to post it there, but I realized that I made the app for iOS only. I hit the next wall: how do I get this on Android without learning a whole new language and rewriting everything? Enter Skip. The idea of transpiling my Swift code directly into Kotlin felt like magic. So I started doing that with a little bit of help from AI and some YouTube tutorials since I had never done this before. This is gonna be so easy, like what do app developers
Proof of the amount of bugs at one point
even groan about…..?
5 days later, with minimum sleep in my system. Let’s just say transpilation was not a smooth ride. Errors at every turn, red everywhere. I had to debug 500 bugs in ONE RUN. So I turned to Reddit, like any sane person to ask people for any other way to solve my problem. One person there told me that “You’re doing it the wrong way. You are supposed to make an app for Android first and then iOS”. Are we serious right now ? Instead of turning back, I pushed through the stack trace, fixed the type mismatches (like ensuring float literals instead of integers in ternary offset calculations), and stabilized the build pipeline.
Me using Gemini 3.1 Pro and Claude Sonnet 5 with Medium effort
My biggest roadblocks weren’t just Swift syntax; they were the growing pains of bridging two completely different ecosystems. I spent hours wrestling with the underlying mechanics such as realizing macOS privacy protections on my Desktop were quietly blocking Gradle from syncing and understanding that Skip’s generated files are ephemeral and require rebuilding the Xcode project to keep Android Studio in the loop. I mostly did it on my own, but since I was participating in a hackathon, i had to speed things up, so I used the free version of Claude, Pro Gemini, and Spotify to help me break through these faster.
So, currently I am working on making the Android app more stable, clean, and useful. It was a struggle, but solving those platform-specific quirks taught me more about cross-platform development than any tutorial ever could.
You can check out all my code on GitHub here (it gets regularly updated)
