We Built a Vacation App in Two Days
We Built a Vacation App in Two Days
Because ordinary maps did not understand our snack strategy, our e-bikes, or the urgent civic importance of finding patios, rooftops, barbecue, coffee, parks, and backup supplies without turning the day into a spreadsheet with handlebars.
There’s a Point in Every Trip Where the Map Stops Being Enough
There’s a moment on a trip when you realize the regular travel tools are not wrong exactly. They are just not living your life.
They do not know you arrived by train. They do not know your folding e-bikes are now part transportation, part personality test, part rolling luggage experiment. They do not know you are staying in Soulard, trying to decide between barbecue, rooftops, parks, coffee, hidden gems, and whether a ride is “easy” or “please verify this route before we become local news.”
And they definitely do not know that Kellie and I are trying to turn St. Louis into a game board, with snacks, stops, rides, and little wins along the way.
So we did the normal thing.
We built an app.
Not a fake app. Not a “someday” idea. A real phone-ready travel helper. The kind we could open on our Android home screen while riding around town, checking saved stops, filtering by distance, and deciding whether the next adventure should involve pizza, a park, a rooftop, or basic survival supplies.
The Problem Was Not “Where Should We Go?”
The real problem was sharper than that.
We needed a tool that understood our actual trip. Not a generic tourist list. Not a pile of sponsored suggestions wearing a fake mustache. Not a map that throws fifty dots on a screen and says, “Good luck, traveler. May your battery hold and your knees forgive you.”
We needed a local-first St. Louis finder built around how we were moving: folding e-bikes, short hops, daylight rides, walkable Soulard stops, and the occasional “worth the ride” adventure.
The app had to answer practical questions fast:
That is when the project changed from “cute side task” to “wait, this is actually useful.”
How We Built It, One Small Win at a Time
This was not one giant heroic coding sprint. Nobody wore a cape. No one said “enhance the matrix” while green code rained down a monitor.
It was simpler and more powerful than that: one step, test it, fix it, commit it, move forward.
We Started with a Local App
First, we built a simple browser app. Search, filters, saved list, CSV export, and Google Maps links. It was not glamorous, but it worked. Working beats perfect every single time.
We Built the Database
The app started small, then grew fast. Food, bars, coffee, rooftops, parks, museums, groceries, gas, liquor stores, patios, speakeasy-style spots, and neighborhood anchors. The little helper put on work boots.
We Added Bike Comfort Tags
Distance alone does not tell the truth. Some rides are easy. Some are moderate. Some need a route check. Some are daylight-only. So we gave the app plain-English bike guidance instead of pretending every mile is equal.
We Fixed the Quick Filters
One filter at a time was not enough. Real decisions are layered. Now we can stack filters like “3–7 Miles + Easy Ride + Pizza” or “Patio + Cocktails + Date Night.” Civilization advanced slightly.
We Made It Phone-Ready
The laptop-hosted version was fine for testing. But on a bike, the laptop being required would be, technically speaking, deeply silly. So we published it as a phone-ready PWA through GitHub Pages.
We Skinned It Team Jellie Style
Then we made it look like something we actually wanted to use. Logo. Skyline. Teal. Purple. Sunset colors. Rounded cards. App-style polish. The utility got a jacket that fits.
Try the Team Jellie STL Finder
The public version is live. It works from a phone browser, and on Android you can add it to your home screen like a small app.
Phone steps:
1. Open the app link above.
2. Tap the Chrome menu.
3. Choose “Add to Home screen” or “Install app.”
4. Open it from your phone like a regular app.
5. Find food before morale collapses.It is not a live GPS router. It is a local-first trip helper. Verify hours, routes, weather, traffic, and battery range before riding. The app is smart. It is not psychic. Yet.
What This Says About AI and Coding
This is the part I wish more people understood.
AI did not “magically” build the app while we took a nap and waited for software elves to finish the CSS. We still had to make decisions. We still had to test things. We still had to say, “No, that is not useful enough,” and “Yes, but can it also find liquor stores?”
Which, honestly, is a valid software requirement.
The power was not replacement. The power was leverage.
We gave the direction. AI and Codex handled the heavy lifting. We tested. We corrected. We pushed forward. The app got better because we kept treating it like a real tool, not a novelty trick.
That is the practical revolution: a normal person with a clear need can now build a useful tool around their own life. Not someday. Not after a four-year degree. Not after memorizing a keyboard shortcut invented by a committee of raccoons. Now.
This matters for creators, travelers, bloggers, small businesses, retirees, families, hobbyists, and anyone who has ever said, “I wish there was an app for this,” while staring at their phone like it owed them rent.
Creator Desk Essentials That Helped Build the Workflow
Affiliate Disclosure: This post contains affiliate links. As an Amazon Associate, I earn from qualifying purchases. That helps support Deep Dive AI and Team Jellie Adventure Corp at no extra cost to you.
These are practical creator tools for writing, coding, editing, and keeping a small AI workflow from becoming a pile of digital laundry.
Logitech MX Keys S
Slim, quiet, reliable keys with smart backlighting—my default typing surface for long writing and coding sessions.
Check price →Logitech MX Master 3S
Comfort sculpted, smooth scrolling, and multi-device switching that keeps the desk from becoming a wrist-shaped complaint department.
See details →Elgato Stream Deck +
Physical knobs and keys for macros, audio, scene controls, and the satisfying feeling that the desk has buttons for a reason.
View on Amazon →BenQ ScreenBar Halo 2
Even monitor lighting without glare, so the late-night writing, debugging, and “why is this cached?” investigations stay readable.
Buy now →Anker USB-C Hub 7-in-1
The little port rescue squad for modern laptops that somehow forgot people still use things.
Get the hub →Product fit: these are creator-workflow items, not travel necessities. They support the AI/coding/content-production side of the project.
Listen While We Build, Ride, and Overthink the Snack Map
Every good build needs a soundtrack. Three blues albums below. Hit play, open YouTube, or let the Delta mood supervise your next app idea.
The Real Lesson: Make the Tool Fit the Life
Most software feels like it was built for a person who does not exist.
This app was built for two people who very much do exist: us, on e-bikes, in St. Louis, looking for food, views, parks, supplies, and the next “this might be fun” decision.
That is why this project matters beyond the trip. It shows what AI-assisted coding can do when the human stays in charge of taste, context, and purpose.
Start Useful
The first version does not need to be beautiful. It needs to work.
Test in Real Life
The best feature requests come from actually using the thing.
Make It Yours
Brand, voice, design, and humor turn a utility into a companion.
We did not set out to become app developers.
We set out to make our trip easier.
Then the app started working. The phone icon appeared. The filters stacked. The saved list made sense. The skyline looked right. The snack strategy had infrastructure.
And there it was: a tiny piece of custom software built around a real life.
That is the door AI opens.
Tags: Team Jellie, Deep Dive AI, AI coding, St. Louis travel, e-bike travel, folding e-bikes, travel app, GitHub Pages, PWA, AI workflow, Blogger, creator tools, vacation planning
Comments
Post a Comment