Out of nowhere, three random people stormed into my room with a hackathon idea. I had no clue that these strangers would become the best team I’d ever worked with. But something about their pitch (which I’ll reveal soon 😉) got me on board—along with one more teammate.
Meet the squad:
🔥 Vedant – ML engineer, made machines smarter while we Googled syntax.
🔥 Saptarishi – ML developer, because one AI geek wasn’t enough.
🔥 Shaurya – Frontend dev + our “market research guy” (whatever that meant).
🔥 Neharika – Frontend dev, made sure our UI didn’t look like 2005.
🔥 Me – Backend dev, keeping things running under the hood.
Team? Check. Idea? Oh, you’ll want to stick around for this. 😏
And just like that, Team Neural Ninja was born. Cool name, right? We thought so too.
Now, the idea—brace yourself. Ever thought about switching to renewable energy, specifically solar? Sounds great, but here’s the catch: solar energy isn’t the same everywhere in India. Factors like location, panel angle, subsidies—it’s a mess.
So, we decided to fix that. Our platform would let users enter their latitude and longitude and—boom!—get everything they needed: energy generation estimates, power consumption insights, ideal panel angles, and even available subsidies.
Oh, and the best part? There was no dedicated marketplace to buy solar panels for homes and small offices. So, we planned to build that too.
The idea was solid. Now, it was time to bring it to life. And that’s when things got real. 🔥
We shot our shot and applied for our very first hackathon—a national-level competition by SAP. Perfectly aligned with our idea. Sounds cool, right? Well, there was just one tiny problem...
We had zero clue how to actually build it. Sure, we knew bits and pieces of our respective tech stacks, but turning our vision into reality? That was a different beast.
Luckily, the first round was just an idea round. Vedant whipped up a basic ML model—functional, but not quite there. Neharika threw together a frontend, which, of course, wasn’t integrated with anything. Time was slipping away, and doubts started creeping in. Would we even survive this round?
Then came the plot twist—the regional round for Chennai was happening on our campus. Sweet! Or so we thought…
Walking into the arena, we were not ready for what we saw—100 teams. The room was buzzing with competition, and to top it off, there were 10 evaluation tables. Our team? Yeah, we got assigned the 10th one. Meaning? We were dead last.
Cue the panic. Going last meant two things:
1️⃣ We’d have to wait forever for our turn.
2️⃣ By the time we got there, the judges would be exhausted.
Enter Shaurya, the game-changer.
I kicked off the pitch, Vedant explained the ML part, and then came our savior—Shaurya—with an intense market analysis. His B2C model completely won over the judges, and for the first time, we had hope. And guess what? We made it past the first round!
But there was no time to celebrate—the second round was the very next day. This time, it was online, and the jury was from the company itself. We made improvements, everything was going smoothly... until we messed up.
Since time was tight, we joined from multiple laptops. But the audio started echoing like a bad horror movie. Our genius fix? Mute one laptop while speaking from another. The only problem? We didn’t hear the jury asking us to stop. We ran five minutes over time.
The moment we realized it, our hearts sank. We thought, That’s it. We’re out. Disappointed, we started looking ahead to the next hackathon…
Guess what? We made it to the final round in Coimbatore! Our first hackathon trip, and we were beyond stoked. The cherry on top? Only two teams from VIT had made it this far.
But then reality hit—the work was doubled. We had a prototype to show early morning and barely anything was done. The ML model? Initially based on a different dataset, but we needed to switch it up, and of course, our accuracy dropped to 60%. The pressure was on Vedant and Saptarishi. They were neck-deep in complex models, trying to up the accuracy, but with exponential complexity came the demand for more RAM than we could handle.
Oh, and did I mention? Our backend was in JavaScript, and the ML model was in Python. Trying to make them talk to each other felt like trying to connect two different languages. The frontend? 70% incomplete. Time was ticking, and we had a train to catch.
That’s when the magic happened—the train-a-thon. The whole team packed into this late-night train, everyone else sleeping, while we geeks spit out random tech jargon, trying to solve this mess. Our fellow passengers? They were glued to us, probably wondering if we were hacking into the Matrix. One generous aunty even came over and wished us luck!
By morning, we had 80% of the solution ready.
Here’s what went down:
🚀 Vedant and Saptarishi cranked the accuracy up to 91% using statistical inference and descriptive analysis.
🚀 Shaurya built the logic to analyze ML data and provide key insights to the user.
🚀 I wrapped the ML model with a fastify backend and integrated it with our main backend.
🚀 Neharika and I finished the frontend-backend integration.
We had our solution—barely in time, but it was ours. And the best part? We were ready to show it off. 💥
The big moment came—we had 10 minutes to present. And guess what? I totally froze. I mixed up my teammates' roles and stumbled through it like a deer caught in headlights. But then, I regained my groove and somehow managed to pull it back together.
What happened next? Pure magic. The judges were so engaged that those 10 minutes stretched to 40. They loved our idea and even gave us some amazing feedback. It felt like a win, honestly.
The next day, the results came out. We didn’t win. But you know what? We were beyond happy. We’d made it this far, learned so much, and gained experiences that no victory could replace. After a whole day of travel in Coimbatore, we headed back to the pavilion—tired, but proud.
And that’s how it all went down—the journey of Team Neural Ninja. 😎🔥
But hold up—this wasn’t the end of our story. We didn’t stop there. We applied for different hackathons, each with a new idea, and everything was going well. But no matter what, our first project—our baby—still demanded a proper closure.
So, we applied for a 24-hour hackathon, and here’s where things got wild.
When I opened the codebase, I had one thought: What. The. Heck. It was a mess—my code was a disaster. But after some serious optimization, I finally got it running smoothly.
Next, we tried to deploy it all fancy-like. Spoiler alert: It didn’t work. We had just 5 minutes left, and we nailed it anyway.
Two days later, the results came out. And guess who bagged first place? Team Neural Ninja. 🎉 We finally gave our baby the closure it deserved. 💥
And that, my friends, is how we went from hackathon noobs to champions.