The journey to shipping SplitMyExpenses on iOS & Android app stores
The mobile apps for SplitMyExpenses are finally out for iOS and Android!! This has been a monumental effort for me. It all started in March 2024 when I made the leap of faith to create my Expo React Native codebase. After so many SplitMyExpenses users asking me: "Where is the mobile app?" I'm happy to announce that after 652 commits, a lot of AI coding, and plenty of pain, sweat, and tears, the mobile apps are live and able to be downloaded right now. I'd love to share the journey and the choices I made along the way.
Download on the App Store
Get it on Google Play
The SplitMyExpenses origin story
I started SplitMyExpenses back in 2020 as a side project in Laravel (PHP) and Vue.js. After working on it extensively in my free time and feeling the undeniable urge to work for myself, I quit my job at Meta after 2 years and released version 1.0 in June 2023.
The reason I built the web app first comes down to practicality and origin story. If you know anything about mobile apps, deploying to the app stores is a HUGE pain. It's heavily regulated. I can commit and push to deploy my web codebase anytime, but for mobile apps, that's sadly not the case. So even though I was (and still am) a mobile developer in my 9-5 day job, building iOS apps for 8+ years, I chose to start with web and did that intentionally. It was also a side project, I didn't want to go full complexity on day one, so web app was way way easier and modifiable.
From personal tool to Splitwise competitor, shipping v2.0
The original version 1.0 was simple: your own personal contact book where all contacts were Venmo users.
Then all my v1.0 users started asking, "Why don't you build a Splitwise competitor? Why not add groups with shared debts where users can collaborate?"
Honestly, I dragged my feet for quite a few months. This wasn't my original vision. I'd built a small, simple tool for myself. But over time, when enough people tell you something, you start to think maybe I'm wrong and there is opportunity there.
At the end of 2023, I released SplitMyExpenses 2.0 after a two-month painstaking rewrite, breaking the original personal-view structure and bringing it to a group-based point of view (literally all DB tables changed and 90% of the UI as well). The feedback was great, phew! The website user base was growing organically month over month, and I was adding features left and right: activity log, recurring expenses, and other game-changing features critical to competing with Splitwise. Yes this was before the advent of Claude Code and the likes, so hand writing it all!
"Where's the mobile app?" (x100 users)
But the same question kept coming up: "Where's the mobile app?" "Is there an iOS app? I couldn't find it on the App Store."
As someone who has built mobile apps their entire professional career, my honest thought was: I don't want to do that. Why would I spend hundreds, if not thousands, of hours building a mobile app when I already have a web app that works just fine? I acutely know the amount of effort it would be and it was a big time commitment now, and in the future.
So, I did what any founder would do, I doubled down on making the mobile web experience first-class, aka spreading my vision. I spent a ton of effort on responsive design: adjusting layouts for small screens, even adding a Chrome manifest for PWA support. I was saying, "I hear you about the mobile app, but let me get as close as I can using the native browser APIs/availabilities."
But people kept telling me, "I'm not going to use this only because there's no mobile app." Eventually, it wore me down. I was frustrated, so were they, well well well... Ok fine, I'll do it. Can't hurt right?!
The tech decision: Native vs React Native
After I mentally decided, F it, I'm going to do it, the question became: what do I pick? Swift native? Kotlin native? React Native? I'm a one-person operation, just me, myself, and I coding these websites and apps. How do I move fast, build something that won't take five years, and still ship high-quality software?
I chose React Native. This was a new learning experience for me as a Swift/Objective-C engineer. I'd been pretty full-stack since building web apps with Laravel and Vue.js for the past 10 years and iOS for the past 8, but React Native was my first time bringing TypeScript to mobile and dealing with React. Since I had a strong Vue background, it wasn't too hard to pick up and with AI, even easier!
One thing I learned quickly about React Native and Expo: there are so many packages out there, and some are critical to making your app high quality. A high-quality React Native app is essentially a collection of high-quality packages stitched together—which is both a blessing and a curse. I'd often choose a library, think "there's got to be something better," and then months later find it and do a migration. Thank god for AI tools, because writing those migrations by hand would be brutal.
Example of just a few:
- TanStack DB
- TanStack Query
- MMKV
- Zustand
- react-native-keyboard-controller
- and so many more good ones...
The Long Road: April 2024 to November 2025
I worked on the app from April 2024 to about June 2024. June was significant for me personally. That was my one-year mark of running my own company: VuxByte. I'd told myself if I didn't see a long term future after one year, I'd go find a job again, no questions asked. Unfortunately, the financials weren't there at the time. SplitMyExpenses was the primary product, and while I'd also released Chatty Butler on the iOS App Store, the AI chat competition was fierce and difficult to pull ahead without being closer to model creation.
So I paused the in progress React Native app completely and started doing full time interviewing instead. I ultimately signed a job offer at Snapchat and pretty much didn't touch the mobile app for months during the interviewing process.
I picked it back up in September 2024 after signing my job offer, worked on it a bit, then barely touched it through the end of 2024 and early 2025. Looking at my commit history: 165 commits in 2024, mostly sporadic and not consistent, it was a back burner project at the time.
Enter Claude Code
Everything changed in late April/May 2025. This is when Claude Code came out, and I went absolutely ham! I've paid Anthropic $200/mo for the Max plan ever since, and boy has it been worth it.
As a professional software developer, I know how to write good code—but it takes time. When I first used Claude Code with their Opus model, I was shocked at how productive I became. The code quality was excellent. I was hooked. I shared about how I loved it on my podcast: Breakeven Brothers I host with my brother Bennett. It was definitely a moment for me, one I'll never forget.
If you look at my commit history from that point onward, it's incredibly consistent: commits every other day, features being completed constantly, ported over from web. From late April 2025 through November 2025, I pushed over 500 commits, all powered by AI-assisted development. To be fair, I used a mix of Claude Code and Codex (since if you are on AI Twitter, you know the whole Claude Code drama), but the majority were Claude Code.
I'll be honest: the SplitMyExpenses mobile app is probably the biggest AI code generation project I have in production. The web app was written entirely by hand in the "old days", then I added AI on top, so maybe 75% handwritten, 25% AI-generated. But the mobile app? After the initial ~150 manual commits, the remaining 500+ were almost entirely AI-assisted. I'd describe the problem/feature/flow in great detail, test the resulting flow post code generation, review the code, and commit it.
┌────────────┬───────────────────────────────────┬────────────┬────────────┬───────────────┬────────────────┬────────────────┬─────────────┐ │ Date │ Models │ Input │ Output │ Cache Create │ Cache Read │ Total Tokens │ Cost (USD) │ ├────────────┼───────────────────────────────────┼────────────┼────────────┼───────────────┼────────────────┼────────────────┼─────────────┤ │ Total │ │ 2,723,574 │ 2,177,510 │ 83,074,004 │ 1,041,567,691 │ 1,129,542,779 │ $711.71 │ └────────────┴───────────────────────────────────┴────────────┴────────────┴───────────────┴────────────────┴────────────────┴─────────────┘
The finish line
The iOS app launched in mid-November 2025. I didn't have to fight the App Store too much since I'd fought (and learned) those battles with Chatty Butler, which I've written about separately. It was a monumental day. I marked it done in my user feedback tracker and moved on to Android. Yesterday, Saturday, November 29th, I released the Android app. It's now live on both platforms. The satisfaction and relief are immense. From the first commit on March 13th, 2024, to today, this journey has been soooo long, but it's done!
What I learned this time around
1. Listen to your customers, they will help guide you
I often shook my head and said "this is my vision", but that was a mistake. If I'd started the mobile app earlier, it would have been less painful since I wouldn't have had to recreate so many web features on mobile. It probably would have been smoother, but it's nice to let your users guide you, and this mobile app wasn't a non-obvious idea, but when they beat the drum about "where is the mobile app" enough times it keeps you up at night, you start to think, yeah yeah I should do this, the vision is changing and I'm all for it now.
2. AI coding is absolutely incredible
I would not have shipped this app for another 6+ months without Claude Code at my disposal (FWIW I have been subbed to the $200/mo Max plan ever since May 2025). I'll say that ten times over and sing the praises of Anthropic, because it is truly incredible software. A great frontier agentic coding model with a great agent harness CLI. Building features became more about describing the right problem and tying architecture together, rather than manually typing out code. I talk to my computer a lot now to say the least. My talking speed = my shipping speed! There is a whole lot of nuance to this which I might write about in a future blog post, but overall it's incredible at the rate you can pump out high quality software if you know what you are doing.
3. It feels amazing to finish a project
The web app existed "alone" from June 2023 to November 2025. Over two years with no companion first class mobile app. Now I finally have a full-featured product across both ALL platforms: web, iOS, and Android. Mobile will serve users who split expenses on the go and bring a first class experience to SplitMyExpenses users in the palm of their hands. I can't tell you how many "side projects" exist in my graveyard, but seeing a project through to completion is a rare achievement and a huge relief.
Looking forward on SME
There's still so much to do: marketing, communication, SEO, ASO, paywalls, you name it. But right now, I'm taking a moment to celebrate this win before moving forward. I'm incredibly proud of how far I've gotten, and I'm excited to share this milestone, the journey, and the choices I made along the way.
Try it out now!
Ready to split expenses on the go? Download SplitMyExpenses for free!
- Download on the App Store (iPhone, iPad, Mac)
- Get it on Google Play (Android)
Both apps are free to download with all core features included. I'd love to hear what you think!