The Hackathon Bond

Jasmeet and I have always loved hackathons. Back in our college days, we collectively participated in over 20 hackathons. For me, they helped shape my early interest in Android development; for Jasmeet, it was robotics.

In the beginning, we were just figuring out what we enjoyed — picking up new tools and building whatever felt exciting. Over time, hackathons became less about winning and more about jamming with friends and building cool stuff.

Some memorable ones:

1. Knocky at NIT Surat — we built a table knock detector that triggered actions according to number of knocks. (Team Snorlax). Me, Akshay and Gursehaj.
2. Hack DTU — an app that tracked heart rate using just a phone camera (Team CodeRage). Me, Manan, Gurpreet and Sahil.

Jasmeet had his own set of wild builds with his crew in college. After college, though, life took over and hackathons paused. But the grind, the chaos, the late-night debugging — we missed all of it. That’s when we started hunting for hackathons we could still join as working professionals.

NxEvos Hackathon and GARUD

HackerEarth has always been our go-to spot for hackathons — even back in college. These days, though, we’re more selective. Our basic rule: if we’ve already won $1,000 somewhere, we won’t join another one offering less. We usually look for events with at least $1,000 in prizes — but it’s not just about the money. The theme matters too. We skip blockchain stuff (just not our thing) and go for challenges aligned with what we enjoy building.

That’s how we found NxEVOS Hackathon — first prize: $5,000. The challenge? Build a real-world solution using the NxEVOS video cloud platform. Their platform helps you process video streams, build dashboards, and even supports plugins and cloud integrations — basically a developer toolkit for anything video-based.

Jasmeet has been working in robotics and computer vision, so this was aligned with his interest. We started brainstorming on a couch — literally — and landed on an idea inspired by my trip to Yellowstone the year before. As a tourist, finding animals like wolves or bears was tough. And I realized it’s not just tourists who struggle — park rangers also need real-time info to protect wildlife. That’s when GARUD was born. Stands for Geo-spatial Animal Reckon User Dashboard. If you like the name, Jasmeet is the one good with naming.

The idea: A video-powered animal tracking system with real use cases for:

• Tourists: See where animals were last spotted.
• Researchers: Access long-term animal location data to study behavior and biodiversity.
• Park Rangers: Track movement to detect threats or help injured animals.

GARUD had multiple parts:

• A frontend to show animal locations on a map.
• A backend to serve location data.
• And the core CV engine, powered by YOLOv8, to process camera feeds and detect animals.

Dividing Work and Building GARUD

We knew what we were signing up for — full-time jobs by day, hacking by night (and weekends). So we played to our strengths:

• Jasmeet focused on integrating video streams using the NxEVOS platform, running YOLOv8 detection, and integration of a local frontend for testing.
• I handled the backend to expose detected locations, the Next.js frontend for showing animal movements on a map, and anything involving docs or presentations.

Because we divided scope clearly, development moved in parallel. But it wasn’t as smooth as it sounds. NxEVOS was a new platform, and the documentation had gaps. Jasmeet especially had to dig around, experiment, and often troubleshoot through trial and error.

Still, we kept grinding. I remember nights where one of us would be awake wrapping up pieces of our part. On weekends, we’d sit together on the couch, put on the Spider-Verse soundtrack (our go-to), and push through whatever was left.

Submission and Announcements

Submissions always end in a rush for us. We could finish a week early, but for some reason, we always take it to the last minute. I usually prioritize the README and presentation — they’re the first impression, like a book cover. If it doesn’t catch attention, nobody dives in.

We wrapped things up and submitted. [Check out the repo and demo video here.]

After that, came the waiting game. We have a habit of checking the hackathon page every few days — especially the discussions tab. It’s fun seeing other impatient devs like us 😂

Then came the morning it happened. Jasmeet was getting ready. I had just woken up and casually checked HackerEarth. And there it was — our team name. For a moment, I sat silently, soaking it in. Then I got up, hugged my mom and Jasmeet, and told them:

We won. 3000 dollars.

Totally worth it 🎉