I'm a UX/UI designer by trade. I know how to make things look good. I know how to make things feel right. I know how to obsess over a button radius and whether a font is doing its job. I know some front-end CSS. Margins, spacing, the good old !important; when something won't cooperate. Designer stuff.
What I don't know is the backend. The logic, the databases, the server stuff, the part where data actually goes somewhere and comes back. That world was completely foreign to me.
And yet, somehow, I built an app. A real one. At a real URL. That real people can open on their phones right now and check it out as demo prototype.
Here's exactly how it happened and I'm not leaving out the embarrassing parts.
Step 1: I started where every designer starts. Figma.
I opened a blank frame and started designing. No plan, no dev hired, no business model. Just a vision and a design tool I already knew.
The homepage came together pretty quickly. It wasn't polished, more of a high-fidelity wireframe with placeholder content, but the bones were there. The logo was on point. The colors felt right. The typography was doing what typography is supposed to do. I'd say it was about 75% done, which in designer math means it was about 50% done.
I also set up a Shopify account. Thought I'd sell Uplyft merch, print on demand, coffee mugs, the whole thing. Ended up connecting Printfy with Etsy instead, and sold some mugs and got zero sales on the SVG packs. Shopify became a landing page with an email capture form.
Nobody signed up.
I shared it everywhere. Nobody signed up.
I kept sharing it. Still nobody. It’s still there ready for signups.
Step 2: I heard about Replit.
Someone told me it was a platform where non-developers could build apps. I was skeptical and curious at the same time, which is probably the right emotional state for a designer trying to become a founder.
Since I couldn't connect Figma directly to Replit at the time, I did what any reasonable person would do. I took screenshots of my designs. Then I went to ChatGPT and wrote the most detailed prompt I could: features, functionality, target audience, monetization ideas, everything. I dropped it in, hit enter, and watched Replit generate an app.
I genuinely thought it was going to be incredible. The anticipation of what it was going to reveal was a good feeling…..but….
It was not incredible.
It only kind of worked. Kind of. It looked basic. Nothing like what I'd designed. I had no idea what the backend code was doing or why certain things weren't working. I spent a couple of months on it, here and there, chipping away making viusal adjustments and some functionality changes.
Then I set it aside. I was bored and frustrated to be honest. This ‘Vibe Coding’ thing wasn’t vibing.
Step 3: I found Claude.
Not as a chatbot to answer questions but as a thinking partner. I described Uplyft the way I would describe it to a co-founder: the vision, the user, the experience, the gaps in what I'd built. Claude helped me think through the architecture known as my tech stack, the strategy, the copy. It helped me see what I was actually building and we fomulated a plan to rework it.
Then Claude Code came out.
This is where things got real. It’s been a game-changer for me.
Claude Code let me work directly inside my project files. I could describe a feature in plain English and watch it get built. Catch a bug, explain what was happening, and have it fixed in minutes. I exported everything out of Replit and brought it into Claude. Claude gave me the prompt to take into Claude Code. And then we rebuilt the app.
That was about a month or so ago. It's been getting better every week. I’m finally getting to a place I actually like. It’s rewarding to see my ideas and changes actually work. I’m happy that it’s actually incredible unlike the first pass with Replit.
Step 4: A live URL. Which required more learning.
Claude suggested I connect the app to a real domain. Simple enough, right?
Not when you've never done it.
I had to create a GitHub account. Then a Vercel account. Then commit all the code to GitHub and connect it to Vercel. Then go to Namecheap, where my domains live, and point everything where it needs to go.
I won't pretend I understood all of it while I was doing it, but I do enough to be dangerous. I’m smart enough to research and ask questions in Claude as well as create prompts to have the agent think like an App Engineer. So, I followed along, asked every dumb question I had, and now app.uplyftlife.com is live and pulling from a GitHub repo that auto-deploys every time I push new code.
I still think that's kind of magical and cool to have it fully automated.
For those who like to geek out: here's the actual tech stack.
If you know what any of this means, genuinely impressive. I'm still figuring most of it out.
What's built:
Figma — Where the designs live. My comfort zone. The one tool I actually knew going in.
Replit — Where the first version was built. Online coding environment, no setup required. Great for getting started, hit a ceiling fast.
Claude and Claude Code — My AI co-founder and the tool that actually rebuilt the app from scratch. I describe what I want in plain English. It builds it. I still don't fully understand how.
GitHub — Version control. Every change to the app gets saved and tracked here. If I break something, I can roll it back. Learned this the hard way.
Vercel — Hosting and deployment. Every time I push code to GitHub, Vercel automatically rebuilds and republishes the live app within about 60 seconds. No manual uploads, no FTP, just push and it happens.
Supabase — The backend database. This is where user data, messages, and app content will live. Open source, free to start, and apparently a big deal in the developer world. I'm taking everyone's word on that.
Namecheap — Where both domains live. uplyftlife.com points to the Shopify marketing site, and app.uplyftlife.com points to the Vercel app. Namecheap is essentially the traffic cop telling the internet where everything goes. I did not know this was a thing six months ago.
Shopify — The marketing site and future storefront at uplyftlife.com. Currently handling the landing page and waitlist.
Beehiiv — What you're reading right now. Newsletter platform. Free to start, easy to use, and the most underrated tool in this entire list for building a real audience.
Total monthly cost at this stage is about $60/mo. Which feels insane given what it does.
What's coming:
React Native, Expo, and NativeWind — The framework that turns this into a real iOS and Android app. Expo handles the builds and over-the-air updates, meaning I can push fixes to your phone without waiting for App Store approval. NativeWind brings Tailwind CSS into React Native so the design language stays consistent. This is how Uplyft gets on the App Store and Google Play. I am told this is the right move. I believe them.
Claude API — The AI brain behind Uplyft's message generation. When you're staring at a blank screen before your kid's game and don't know what to say, the Claude API helps find the right words. This is one of the features I'm most excited about. It's also from the same company whose tools I used to build the whole thing. Full circle.
RevenueCat — Handles subscriptions and in-app purchases across both iOS and Android without me having to figure out Apple and Google's billing systems separately. Free until I hit $2,500 in monthly revenue. I look forward to that being a problem I actually have.
Stripe — Payment processing for the web side. The thing that actually moves money. Everyone uses Stripe. Now I will too.
Apple Developer Program — $99 a year to get on the App Store, run TestFlight beta testing, and enable Apple Sign-In. One of the few things in this list with a price tag. Worth every penny when the app is ready.
Google Play Console — $25 one time to distribute on Android. Twenty five dollars. That's it. Still kind of amazed by that.
Google Cloud Console — Handles Google Sign-In authentication. Free. I had no idea this was even a separate thing until someone told me. Now I have an account.
Push notifications via Expo — So Uplyft can tap your shoulder at exactly the right moment. Game day morning. An hour before the audition. The thing that makes scheduled messaging actually work the way it's supposed to. Core to the product.
Stripe — Payment processing for the web. The thing that actually moves money.
NFC chip integration — This one's my favorite and the one people look at me the most sideways about. Uplyft will have a physical bracelet embedded with an NFC chip. Your kid taps their phone to it and gets their personalized message, their playlist, their pre-game mindset content. No app hunting, no notifications to swipe. Just tap and go. It's also, honestly, a pretty great investor demo.
Phase 2 and beyond: Twilio for SMS messaging, Spotify integration for motivational playlists, and GameChanger ans TeamSnap for syncing your kid's game schedule directly into Uplyft. These aren't needed yet. But they're on the map. I don’t know if GameChanger has an API, but if it does, Claude and I will implement it.
That last one is still a ways out. But it's on the roadmap, and I think about it every time I watch my kids lace up their cleats.
So, What's next?
Getting the React Native build onto the App Store. Then Google Play. (Once the prototype is working and I have some user testing of course. 😀 )
I have almost no idea how to do any of that yet. But that's been true at every step of this process, and somehow I'm still moving forward.
I'm just a UX/UI designer who knows his way around CSS. Winging the rest of it with really good tools and an AI partner that doesn't judge me for not knowing things….at least to my face. He probably talks to his AI buddies at his watering hole.
So, What's next, next?
The app is live. You can visit it right now at app.uplyftlife.com. It looks real, because it is real.
But it's a prototype. A working one, but still a prototype. You can't create an account yet. You can't add family members or send messages. The core features that make Uplyft actually work are still being built.
What I need right now isn't paying customers. It's honest eyeballs.
Click around. Look at the screens. Tell me what's confusing, what's missing, and what made you think "okay, I actually get what this is."
Hit reply with whatever's on your mind. One sentence is enough. You don't have to be a designer or a developer to give useful feedback. You just have to be a parent, or someone who has ever wanted to say the right thing to someone they love at exactly the right moment.
That's the whole product. And it's almost ready….50% in designer math.
When it’s launched, sign up for the Founding Member offer: $4.99/month locked in forever for the first 100 people, before it moves to $7.99/month Standard or $12.99/month for the full Family Plan.
There will be a Free option too, but it will be super limited.
Thanks for reading. Hopefully this inspires you to take action and build something…..even if you have no idea how to do it.
All the best!
Brad
