I Joined a Four-Person Team at an AI Hackathon. It Was So Different from Building Something Alone!

On October 17, 2026, I joined the Tokyo Agent Hackathon in Tokyo!

This time, I ended up working with a team to build an AI app.

I've participated in AI hackathons before, but somehow, every event brings a new discovery.

And the biggest thing I learned this time?

Building something on your own and building one thing together as a team are much more different than I expected!

So Many People Were Building Solo!

Like the other international AI hackathons I've attended, this event was conducted almost entirely in English. There were occasional explanations in Japanese, but English was definitely the main language.

One thing I found funny was how often I heard phrases like "In Japan" and "Japanese something-or-other."

Well, of course! But it was still funny!

Another surprise was how many people were working as one-person teams.

I had come hoping to join a team, so I was planning to find people to work with at the event.

But there weren't nearly as many people looking to form teams on the day as I'd expected.

I really wanted to build something with other people, so I asked to join a team.

And in the end, we became a team of four!

Working with people I'd just met, trying to build an app together in a limited amount of time.

That's one of the things that makes hackathons so interesting!

It Didn't Have to Be My Idea

I had brought an idea for an AI app that would help manage shopping and luggage while traveling overseas.

The idea came from my own experiences: buying too many souvenirs, worrying about airline baggage limits, and trying to keep track of everything.

Wouldn't it be useful to have an app that could manage your purchases and luggage weight from before your trip until you returned home?

That's what I'd been thinking.

But after discussing our ideas as a team, we decided to build something else.

And I was completely fine with that!

Of course, bringing my own idea to life would have been fun. But this time, I also wanted to experience building something with people I'd never met, even if it was a topic I wouldn't normally choose myself.

Once we'd decided what to build, it was time to get started!

First, Research. Then, UI/UX!

My first task was researching information that could be useful for the app.

I looked for publicly available documents and data, trying to figure out what information we could access and how it might be used in the app.

I'm usually busy building my own apps, so doing research for someone else's idea felt surprisingly fresh.

I also discovered that even when public data exists, getting exactly the information you need isn't necessarily easy.

"Data is available" and "data is ready to use in an app" are two very different things.

That was an interesting lesson.

Later, I also got involved in the UI/UX work.

At AI hackathons, it's easy to focus on what the AI itself can do. But the screens people actually interact with are important too.

This time, I thought it might be fun to make the experience a little more friendly and approachable.

Somehow, I Ended Up Creating a Character!

We decided to add a chatbot, so I started working on a little character for it.

I wanted something that could help explain the app's features and make it easier for users to ask questions.

The first character Claude Code created was just way too cute. Haha!

I wasn't sure it really matched the theme of our app, so I wanted to change it.

Hmm. Something a little more Japanese might be nice.

That's how I ended up creating a character inspired by a Shiba Inu.

I did wonder if it was a little too similar to Hachiko, the famous dog associated with Shibuya!

Then I adjusted the size, making the character a little bigger.

But there was another thing bothering me.

What if people thought it was just a picture and didn't realize it was a chatbot?

The way I'd built it, the question input box appeared only after someone clicked the character.

So I added a speech bubble saying, "Ask me anything!"

Maybe that would make it easier for people to realize they could click it?

It's a small adjustment, but when you start thinking about the people who'll actually use your app, these little details suddenly feel important.

Of course, just because you've built something doesn't mean it's ready to become part of the finished app.

Still, being able to experiment and turn an idea into something visible is so much fun!

Then I Realized: The Hard Part of Team Development Comes at the End

This was probably my biggest discovery of the day.

When four people divide up the work, they can accomplish more than one person working alone.

Someone can research, someone can build features, and someone can focus on the UI.

Being able to work on different parts of a project is one of the great things about having a team.

But...

Combining all those separate pieces into one working app is a completely different challenge!

Just because something works on your own computer doesn't mean it'll automatically work inside the main app.

How should the interface connect to the features? How should the different parts fit together?

There are so many things to consider.

I've built quite a few apps on my own using AI tools.

When I'm working alone, I can change my mind halfway through and adjust the whole project myself.

But when you're working in a team, you have to think about how your work connects to everyone else's.

I hadn't realized how difficult that could be.

It wasn't about any one person doing something wrong. Building one product together simply comes with its own challenges.

Next time I work on a team project, I'd like to think more carefully about the overall structure and how everything will fit together right from the beginning.

That was a pretty big lesson for me.

Actually, even joining in on something another person had already started building was challenging.

I wasn't sure where to begin or how I was supposed to work on it.

In the end, I couldn't figure out how to collaboratively edit the project that had been built with Claude Code. So I got a copy and started working from there.

I wonder if this kind of thing is no big deal for people with stronger technical skills??

I'm gradually getting used to building apps on my own with AI.

But this experience made me realize that working on software together with other people requires a different set of skills.

Time for the Pitches! And So Many Interesting Projects

When the building time was over, it was time for the presentations.

Each team introduced what they'd built in a short amount of time, presenting their app to the other participants and judges.

Our team didn't make it into the top five.

But watching the other projects was an experience in itself!

One that really stood out to me was a wildfire detection and monitoring app using data from NASA and other sources.

It displayed information on a map, and I found it fascinating to see AI being applied to such a real-world problem.

Another project I loved was an AI music generation app!

Watching it create music while changing instruments and musical styles made me think, "Wow, AI can do that too?!"

There were also projects related to healthcare, culture, security, and career support.

Everyone had been working at the same event, with the same limited amount of time, yet the projects were completely different.

That's another thing I love about hackathons!

We Didn't Win, but I Took Home a Lot

I didn't come away from this hackathon having learned only about building AI apps.

I experienced joining a team with people I'd never met before.

I learned what it feels like to work with an idea that wasn't my own.

To be perfectly honest, the topic we worked on was something I probably never would have chosen myself.

And that's exactly why I learned so many new things.

I also got to try a different role from the ones I usually take on.

And, of course, I discovered how difficult it can be to bring separate pieces of work together into one product.

These are experiences I probably wouldn't have had if I'd just been working alone.

The biggest lesson that stayed with me was this:

Making something work on your own is not the same as making something work together as a team.

Thanks to AI, it's becoming possible to build all kinds of things on your own.

But when you start creating something with other people, a whole new set of challenges—and a whole new kind of fun—appears.

This hackathon turned into another great learning experience!

Next time, I'd love to put what I learned into practice and get better at building things as part of a team.