Beyond the Score: Rewriting Exam Evaluation Systems

It all started when AMD visited our college and introduced an incredible opportunity to win cutting-edge hardware by developing a unique solution. My friends and I were buzzing with excitement, brainstorming ideas that ranged from the wildly ambitious to the surprisingly simple.

One day, a friend of mine, who runs his own startup, shared an intriguing concept: "Problem Founder Fit." It’s all about finding a problem in your life and solving it. That’s when it hit me. I’d always struggled with my exam scores, no matter how much I studied. My scores seemed to plateau, and I couldn’t figure out why.

At the same time, I was diving into a book on KAIZEN, the idea of continuous self-improvement. It dawned on me—how could I improve if I didn’t understand where I was going wrong? This realisation became my driving force.

We’ve all been there—struggling with exams like the SATs or GREs, where every answer has to be spot-on according to study materials. It’s frustrating when generic answers from Google or ChatGPT don’t quite fit the bill.

So, after getting our hands on the hardware, my team and I decided to take a page out of the startup playbook: research first, then prototype. Tried Google forms for research but all i got were responses that didnt provide much insights.

Thats when we decided to push it a step further and talk to our target audience -(Students). Since we were in a college we were able to find a lot of them.

We talked to students (our fellow college mates) and asked them about their biggest exam struggles. The common thread was clear: students wanted to know exactly why they got questions wrong and how they could improve. We asked students from different colleges and high schools send us their views on our software idea.

clips of students giving insights on our idea.

We were excited but also a bit overwhelmed—this was our first dive into AI/ML, and we were total rookies. But instead of letting that stop us, we played to our strengths. We broke the project into smaller tasks, each of us tackling a different part. We explored Hugging Face models and, with some help from our professors, zeroed in on models that could analyse answers and provide meaningful feedback.

And that’s how our prototype came to life:

Link to a simple demo of the project

Our journey from feeling stuck to creating a real solution was driven by personal frustration and a desire to help others. We hope that our project will not only make a difference in students’ exam preparations but also inspire others to find and solve their own problems.