Why We Won
In today's world, hackathons sometimes feel like vibe-coding competitions.
I'll explain why I feel that way in a bit.
So, this was the first time I went to a hackathon.
What I thought it would be, and what I actually did and learned, were completely different.
At the start, I had a pretty clear idea of how we should approach the 24 hours.
I made a plan, mapped out the steps, sketched the architecture, and we got started.
To give you some context:
We split into two groups.
@Nitesh and @Basavaraj worked on the hardware, and @Pratham and
I were on the software side.
And for our project, integration between software and hardware was super crucial.
Most of the others around us were working on pure software.
Our problem statement was about irrigation:
Farming in drought-prone areas.
Now, we knew that what we were building wouldn't directly solve that issue,
but we had a pretty solid approach through polyhouses.
(If you're curious about what exactly we were building and how, check this out: [WEBSITE])
So, now coming back to the "vibe coding" part.
I actually realized it's not quite a thing yet.
And here’s why:
1. We don’t have a reusable, efficient workflow with AI integration on our machines.
Something that could speed things up and help us build anything faster.
2. Even the tools that do integrate AI right now still don’t know enough about themselves.
They can code, yes, but they don’t understand the tech stacks or packages they are running on.
I'll give you an example.
I wanted to test if Cursor or Gemini CLI could build a working team of agents that could coordinate and use tools efficiently.
Both failed.
I gave them multiple prompts, tried various approaches, gave them enough chances,
but they kept failing again and again.
In the end, I had to go back to reading raw documentation and do the coding myself.
So yeah, these were two big lessons I learned about coding and AI workflows.
There were also some powerful non-technical lessons.
Like the importance of deadlines.
I used to read about it in blogs and hear about it on podcasts, but I never actually pushed myself this hard.
We had just 24 hours.
Barely slept 2 to 4 hours.
But we worked like crazy.
Because there was a plan, a vision, and we had to stick to it.
Also, the energy around us, so many people building cool stuff,
That's all the motivation you need. No extra dose required.
And then there was mentorship.
Not in the typical sense, not some professor or tech celebrity.
Just someone who knows the space you are working in.
Around midnight, around 11:30 or so, a senior from the college (@Reva) came to our table.
He was actually part of the organizing team.
He asked what we were working on.
We said, "We are building automation for polyhouses."
And the moment he heard that, his face lit up.
He got super excited.
We sat down in a circle and talked with him for nearly 1.5 to 2 hours.
I even closed my laptop. Didn't touch a single key during that time.
Why?
Because this guy was exactly the person we needed to speak to.
His father is a farmer.
They grow roses, but not in polyhouses.
Still, he knew the problems and the market very well.
He told us everything, actual numbers, prices, labor issues, costs, real-world struggles.
And at the end, he said:
"Bro, I will be your first customer if you build this."
That conversation gave us the one thing we were missing for months:
Clarity.
We had been building something for 2 or 3 months without knowing if it would really help anyone.
We were not sure who would buy it, if it added real value, or if it was just a perfect robotics demo.
That night, that conversation, gave us direction.
After that, I remember thinking:
Even if we don't win this competition, I already got what I came for.
I know I didn’t explain why we won, because @prathma/nitesh/gurugol already did ?Check these posts out