🏆 From Hospital Bed to Hackathon Glory: How We Won 2nd Runner-Up at Datathon 2025 🚀

Every victory has a story, and ours is one of resilience, teamwork, and unwavering determination. Winning 2nd Runner-Up at Datathon 2025 wasn’t just about the prize — it was a culmination of sleepless nights, technical struggles, and a relentless belief that we could make it, no matter the odds.

Here’s our journey, from registration to the podium.

---

The Beginning: A Rocky Start

We registered for the hackathon with excitement and big dreams. The first round was online, and my friend handled most of it while I was unexpectedly in the hospital. Despite the setback, my team pushed through, and we qualified for the offline round. That qualification felt like a beacon of hope — a sign that we had what it takes to compete at a high level.

The Challenge: Cracking the Dataset

The problem statement was intense. We were given three datasets: train, test, and labels. Our task?

➡️ Build a machine learning model to predict item accuracy for a store and generate a complete, correctly formatted data file.

The organizers had strict evaluation metrics, and the competition was fierce. But we were ready to give it everything we had.

The Grind: From 6th Place to the Leaderboard Top

At first, progress was slow. Our initial model accuracy? 0.45.

We refined it. Accuracy climbed to 0.54.

We tried various approaches, experimenting with core ML techniques, hyperparameter tuning, and feature engineering. Our kernel kept crashing. The training sessions would break midway. It was frustrating, but we didn’t quit.

Then, the breakthrough happened. My teammate built a model using TF-IDF and Logistic Regression, pushing our accuracy to a stunning 0.79%. It wasn’t just a marginal improvement — it placed us 2nd on the leaderboard.

I had tried pre-trained models, hoping to leverage their power, but they didn’t click for this problem. Still, every failure taught us something, inching us closer to success.

The Final Push: Fighting Against Time & Crashes

The final stretch was chaotic. Our model's training time was 2.86 minutes, but our kernel was constantly crashing. Every run felt like a gamble — would it finish or break again?

We debugged. We optimized. And somehow, we crossed the finish line with a stable, high-accuracy model that secured our spot as 2nd Runner-Up.

The Takeaway: Believe in the Grind

Winning this prize wasn’t just a validation of our technical skills. It was a reminder that perseverance, teamwork, and belief in oneself can conquer even the toughest obstacles.

If you’re chasing a dream, whether it’s a hackathon win or something bigger — keep going. Fail, learn, adapt, and come back stronger. The win is waiting.

#Hackathon2025

#Datathon2025

#MachineLearning

#Resilience

#Teamwork

#GrowthMindset

#Inspiration