← Video Library · Case Studies

From Idea to App in 8 Hours: Austin's AI Travel Concierge Story

Published February 9, 2025 · 185 views on YouTube

Can someone with no coding experience build an app in a day using AI?

Yes. Austin Distel, who does not know how to code, built a working AI travel concierge app during an 8-hour Gen AI University hackathon by using ChatGPT o1 Pro as his “senior developer” — writing the code, pasting it into his app, and iterating with feedback and screenshots until it worked.

What problem was Austin trying to solve?

Austin, a longtime travel planner, wanted to fix what he saw as a broken travel-booking process: pulling personalized itineraries from scattered notes, email confirmations, Google Sheets, and booking sites. His idea was a marketplace app that takes a travel influencer’s trip history and personalizes it into an itinerary based on a user’s own preferences — something he estimated could save at least 20 hours of planning per trip.

How did he build the MVP over the course of the hackathon?

  1. Hour one — plan and whiteboard. The group mapped out the big idea, whiteboarded possible approaches, and identified which AI tools might fit.
  2. Break the idea into small rocks. Austin broke his AI travel app concept into small, achievable pieces rather than trying to build the whole vision at once — the goal was a minimal viable product (MVP), not something polished or scalable.
  3. Copy-paste travel plans into ChatGPT. He gathered his notes, flight confirmations, and booking details and pasted them into ChatGPT o1 Pro, asking it to generate a clean JSON file for a pretty itinerary.
  4. Paste the generated code into the console. He copied what ChatGPT produced directly into his app’s console to get a first working version.
  5. Iterate on design and logic. When the result didn’t look right, he gave feedback on the logic and design categories, and when code errored out he took a screenshot and told ChatGPT to fix it.
  6. Use Operator to gather bookings automatically. He had OpenAI’s Operator log into Trip Advisor and his email to retrieve his New Year’s Eve trip bookings, saved items, and flight confirmations for Paris, and pull hotel bookings from Booking.com.
  7. Generate a full trip itinerary. By around hour five, better prompting produced SEO-optimized, Apple-like styled code that, once pasted in, generated his complete Paris trip itinerary in one place.
  8. Build a travel persona. Using everything ChatGPT now knew about his past trips, Austin asked it to build a persona of his travel preferences — home airport, seaside 10-day stays, spas, sightseeing, good restaurants — and used it to generate his upcoming one-year anniversary trip itinerary.

What tools did Austin use to build the app?

He used ChatGPT o1 Pro as his coding assistant throughout, and OpenAI’s Operator to browse the web on his behalf — logging into Trip Advisor, his email, and Booking.com to retrieve trip bookings and confirmations without him doing it manually.

What was the outcome?

By the end of the hackathon, Austin had a working app displaying his real itineraries (including Iceland and Paris trips) and had generated a new personalized anniversary trip itinerary based on an AI-built travel persona of himself and his wife’s preferences — all without writing a single line of code himself.

Full video transcript

Austin, what we got going on here on mapping out an AI influencer travel app is that minimal viable product that I can use AI to bridge the gap between what your goals are — for your family trip, your surprise engagement, whatever that experience is — and we’re going to use AI to personalize that itinerary for you. So that’s the goal, let’s see what we can do.

This guy, Austin Distel, built an AI influencer travel app minimum viable product in one day, using AI to code the entire thing for him. He’s not a coder. He built this app from an idea he had in his head for a problem he wanted to solve, and he documented his journey in our one-day Gen AI University hackathon.

First step of the day: I’m going to map out what are the small rocks that could end up leading to a great first MVP experience — that minimal viable product that I can use AI to bridge the gap between what your goals are for your family trip, your surprise engagement, whatever that experience is, and we’re going to use AI to personalize that itinerary for you. So that’s the goal, let’s see what we can do.

Hey guys, so I’m Austin Distel. I was on that early team at Jasper and helped them go from zero to a $1.5 billion valuation in 12 months, getting us to our first 100,000 users. I’ve been using OpenAI since it was just GPT-2, and it was pretty much autocomplete for sentences. I’ve been on a 12-month sabbatical — I took off in December of last year, and over the last 12 months got married, traveled for a bit, moved out here. Now I’ve started to realize over the past year that the way we book travel is broken.

As a travel planner myself, I like to think about my 30 days in Bali and I want to know what are the best restaurants, what hotel should I stay at, based on my own explorations, and so on. I believe that the way I’ve curated my really romantic trips, or curated trips for friends’ bachelor parties, has this standard of just the best, most unique experiences. And I look at travel influencers like The Bucket List Family, or one of my best friends who went to every country in the world — 197 countries. When I talk with him I’m like, “Dude, how did you get to North Korea, what did you do when you were there?” Taking the way that he travels but personalizing it for the way I want to travel — with my wife, or if I do a solo trip — takes a lot of brainpower, but I know AI could be the bridge.

I want to travel like The Bucket List Family, but maybe I want to do it my way. So if I could create a marketplace app that takes somebody’s social media profile with all the trips they’ve been on, and personalizes that into an itinerary for me based on my preferences, that would be magical.

I think it would take at least 20 hours per trip, and I go on a few trips a year, so I think there’s a lot of time saved. I think this is a problem that most people probably face — if you’re somebody who likes investing in travel, this app is going to be amazing for you. And for people who are organized, how can you take all of your confirmations from your email and put them into an itinerary that’s your pamphlet — that book of all your receipts, your flight confirmations, your experiences, your restaurant reservations, all in one spot. At the end of the day, that’s what it is: it’s the AI bridge for content creators and travelers to come together to build the most amazing travel experiences.

Today I’ll be exploring OpenAI’s new Operator system, looking for ways I can send it to profiles, blogs, and other places to curate travel plans based on my preferences. So that’s my big rock of the day.

This is the part one recap of our AI hackathon this past weekend at my house in Austin, Texas, where a group of entrepreneurs got together and used the latest disruptive AI tools to solve a problem in our business or in our life. I think we’re in a golden age of AI right now, and this is a beautiful time to be alive as a visionary, because you have great ideas and you want to turn them into reality.

So I’ve been thinking about this AI travel app for a little while, and over the course of the weekend we took my grand ideas and broke them down into small parts using ChatGPT o1 Pro and the new Operator. We took all of my travel trip plans and turned them into a travel AI concierge app. And look, I don’t know how to code, but I can copy and paste, so I pulled up my senior developer, ChatGPT. I took what it gave me — look at this, it created the code — I put it over here into the app, pasted it, saved it, went to my trips, and now look at this — here’s Iceland right there. I turned this into this over the course of a weekend using ChatGPT.

I’m making this short video series to empower the visionaries, the entrepreneurs, the idea people, to turn their dreams into reality using AI tools, because I believe that greatness is coming soon for you.

This is the part two recap of the AI hackathon we had this past weekend, where a group of entrepreneurs came together to solve a problem or build something with AI over the course of eight hours. Maybe seeing this approach will inspire you to build something too.

In the first hour we took our big idea, made a plan, whiteboarded all the different ways we could approach solving this problem, and figured out what AI tools might be best suited. We then broke that plan down into small rocks. When I was in YC we had this idea of an MVP — a minimal viable product, something so small but that actually solves the problem. It doesn’t need to be pretty, it can be ugly, it doesn’t need to be scalable — it’s just for you, and it’s going to solve a problem. And if you like how it solved the problem, maybe others will too.

So I took my idea of an AI travel app and broke it into its small pieces, and over the course of eight hours I turned things from the backlog, to work in progress, and then eventually done. You would be amazed at how much you can get done on a Saturday with AI tools and a great community around you.

Part three: I’m going to show you the different tools I used — mostly o1 Pro and Operator — and how I turned my itineraries that were scattered all over the place into a travel app. Remember, I’m a creative type, I don’t know how to code, but ChatGPT wrote all of this. Stay tuned for part three.

This is the part three recap of the AI hackathon we had this past weekend, where we used AI to build something awesome in eight hours on the Saturday. So what I’m building here is a travel app — I’m using ChatGPT o1 Pro to code the entire app, to be my senior developer, because I don’t know how to code at all. But my travel plans are all over the place — I take my notes of friends’ recommendations and places to go and things to do, I sometimes organize it in Google Sheets with all my flight information just to try and keep it in one place, I have my reservations like Booking.com — so I need to collect all of these in one place using AI to then create an itinerary. So I’m going to show you how I did that.

Step one: copy and paste all my travel plans. This is the MVP — just copying and pasting text, but in the future we’ll retrieve things automatically. I said, “generate the JSON file to make this a clean and pretty itinerary.” After a few minutes of thinking, I just copy and pasted what it gave me, put it into the console, and bam — I have something. But I’ll be honest, it didn’t look exactly how I wanted it the first time, so I went back and forth. I gave it feedback on the logic, said here’s how I want it designed, here’s the categories we need, and I just voiced my vision to make it better. Sometimes the code would error out, and I’d literally just take a screenshot and say “fix it,” and it did — it would just come back with better code. Stay tuned for part four, where Operator will go to other websites on my behalf.

This is part four of the AI hackathon recap. Now I’m using the ChatGPT Operator to go and browse the internet on my behalf to do tasks. This task I wanted to do was go to Trip Advisor, look for my bookings, and gather all the information about things I booked for my trip to Paris. It logged into Trip Advisor, found my New Year’s Eve trip, gathered all the things I have saved and all the things I booked, and retrieved that list right here. It also went into my email, got my flights from Paris by simply typing “flight confirmation Paris” into the search bar — it retrieved all of that information. I watched it, like a Tesla with no hands, go on its own and find all of my hotels here at Booking.com and then retrieve them.

So around hour five of the hackathon, I got a little bit better at prompting it to create better-looking code. So here below is my trip to Paris with all the notes — I wanted it SEO-optimized, I wanted an Apple-like design experience, and so on — and then it generated the code. I copied and pasted that in, and look at this — it created my entire itinerary for my trip to Paris, all in one place. In the next video, I’m going to show how the AI travel concierge created and generated this trip from scratch based on my personal preferences. Stay tuned.

This is part five, the final piece of the recap of the AI hackathon this past weekend here in Austin, where we built something really cool with AI in under eight hours. My idea was to create an AI travel concierge that would pull in all of my past flights, bookings, and reservations, build trip itineraries, and put it into an app. I don’t know how to code — ChatGPT did all of this work for me. Now that it knows all of my past trips, I wanted to create a persona of who I am to now generate my upcoming one-year anniversary trip.

I asked it to build me a travel persona of who I am, and it nailed it — it understands that I often depart out of Austin, so that’s my home airport, I like seaside 10-day stays, I love spas, I like sightseeing, I like good restaurants. It created this code, which I copied and pasted over here, and check this out — it created my wedding anniversary trip with the flights, and this is so freaking cool. It understands the way my wife and I travel, and it just saved us so many hours of work.

So if you like this kind of thing, if you like tinkering, if you’re an entrepreneur, a visionary, but you need help getting together with a group of people to build something really cool, let me know in the comments the next place we should host an AI hackathon like this. I hope you got inspired, and you build something awesome with AI. If you like this content, let us know what you found helpful and useful, share this and like this so that we know you got value from it, because that’s why we make these videos for YouTube. So let us know, and I will see you in the next video.

Watch on YouTube

Related tutorials