24 hours, partial clarity, and a software visualization idea

I got into the top 3 at the AI Boomi Hackathon.

But this post is not really about the rank.

It is more about what happened in those 24 hours.

Because 12 hours into the hackathon, I was still not fully sure what I was building.

I had started with an idea around agent skills, governance, security, and capability. It sounded interesting in my head. But when I started talking to people, I realized it was too broad for a 24-hour hackathon.

More importantly, I was not close enough to the problem yet.

So I changed direction.

The new idea was around software visualization.

The thought was simple: codebases are hard to understand because they are not static. They keep changing. And most of the time, we consume those changes through files, diffs, and PRs. But a PR is not just a diff.

It changes flows. It touches modules. It shifts the mental model of the system.

So I started building a tool that helps people understand codebases and code changes visually.

A kind of map that lets you see the structure, go deeper into modules, ask questions with context, and understand how a change affects the system around it.

Was the idea fully clear when I started building?

No.

And that was the uncomfortable part.

I went solo. I had to talk to people. I had to explain the idea again and again. I had to take feedback when the idea was not landing. I had to accept that the first version in my head was probably not the one I should build.

Then the building started.

I slept around 2.5 hours in small chunks. Woke up. Checked the build. Fixed something. Changed something. Then again.

Slowly, the product started taking shape.

Rough, but usable.

The submission was also a small adventure. The demo video was not ready until the last few minutes, and I was still thinking about how to explain the product clearly.

Then came the pitch.

I got to present it to Devansh and Pooran, who were amazing jury members. Their questions were sharp and fair. That conversation helped me understand my own product better.

And this is where the real learning came in.

It is great to sharpen your axe before you start working.

But it is equally important to start somewhere.

You cannot get absolute clarity before building. Clarity comes from the loop.

Build a little.

Talk to people.

Get feedback.

Change direction.

Build again.

That loop was the real hackathon for me.

Of course, getting 3rd place felt good. But the bigger thing was how many uncomfortable positions those 24 hours put me in.

Talking to strangers. Explaining a half-formed idea. Being questioned. Changing direction.

Pitching before feeling fully ready.

All of it stretched me.

And I think I needed that.

Thank you to all the amazing people I met there. The inputs, questions, and honest feedback shaped the project a lot.

The whole experience felt really serendipitous.

And I am glad I showed up.