Challenges Are the Way to Grow
Programming a unique application can be challenging, but it always leads to the path of greatness.
BLOG
Abhinav
10/19/20252 min read
Programming a unique application can be challenging, but it always leads to the path of greatness.
Last week, I had a project presentation session at my college where everyone had to showcase their self-made projects. The projects could be in any language — with or without a user interface — as long as they were original ideas.
I created an AI-based researcher that can summarize large research papers, even those with thousands of pages, by splitting them into smaller, processable chunks.
When my turn came, I presented my project to my teacher and explained it in detail. I told her that it uses Perplexity’s API to fetch real-time data, process it according to the user’s query, and even analyze potential research gaps that could be explored in future studies.
She listened carefully and said, “Okay, your UI is perfect, your application performs very well and runs robustly without errors, and it provides a user-friendly experience. But tell me one thing — why should I use your application when I already have Perplexity, which is faster and more reliable? If Perplexity can do everything your app does, why should I choose yours over other AI tools?”
At that moment, I felt a little frustrated. I had put so much effort into building this project, crafting system prompts for researcher-like responses, and ensuring everything worked perfectly. Yet, she said it was no different from Perplexity.
But after some time, I realized she was right. If my application does exactly what other tools do, then what makes it unique? That question made me think deeply.
I started researching the weaknesses and limitations of existing tools to find what could truly differentiate mine. I discovered one major gap: most AI tools treat their answers as 100% accurate, but they are rarely reviewed or cross-verified.
So, I decided to add a Confidence Dashboard to my application — a feature powered by classic machine learning models to compare multiple research papers on the same topic and identify inconsistencies or potential mistakes.
That’s when I understood that challenges may feel frustrating at first, but once we overcome them, we discover our true potential.
Thank you for reading my story. Follow me to get more stories like this every week!
Also, connect with me on:
Instagram — Instagram
LinkedIn — Abhinav Singh
YouTube — Dexterity Coder — YouTube