Why I Built HelloCelpip

I built HelloCelpip because I couldn’t find a realistic, affordable way to practise for CELPIP.

I was preparing for the test myself. The result mattered, but the available options never felt quite right. Tutors were expensive. Practice material was limited. And even when I completed a speaking or writing response, it was difficult to know what was actually wrong or how I could improve it.

I didn’t begin with a grand plan to start a company. I was trying to solve a problem I understood personally.

The product I wanted to use

I wanted realistic questions, useful feedback, and a clear explanation of what would make an answer better. Most importantly, I wanted the evaluation to be honest. A generous score might feel good during practice, but it would not help on test day.

So I started building. The early version was not a complete platform. It was a way to test whether software and AI could give CELPIP candidates feedback that was specific enough to be useful.

As I used it, I kept finding problems. The prompts needed to feel more realistic. Evaluations needed to be stricter. Feedback needed to explain the score rather than simply produce one. Each problem became another feature to build or another system to rebuild.

Becoming the user again

Building the product did not remove the anxiety of taking the real test. If anything, it made me more aware of every weakness in my own answers.

But the strict practice helped. I eventually scored 10 in Listening, 12 in Reading, 12 in Writing, and 11 in Speaking—an average CELPIP level of 11.

That result mattered to me for more than the score itself. It showed me that the kind of preparation I had wanted could work.

From personal tool to first product

Other people had the same problem. They wanted to practise without paying for a tutor every time, but they also wanted more than sample answers and generic advice.

HelloCelpip slowly became a real product. People signed up. Some became paying customers. They asked questions, disagreed with scores, found bugs, requested refunds, and showed me where the product was confusing. That was when I began learning the difference between building software and building a business.

The code was only one part of it. I had to think about pricing, subscriptions, customer support, marketing, evaluation quality, infrastructure, taxes, and dozens of decisions I had never needed to make as an engineer.

I am still learning all of it.

Why I am writing about it

HelloCelpip is my first product. I have made decisions that worked, mistakes that cost time, and changes that only seemed obvious afterward.

I don’t want to turn those experiences into generic founder advice. I want to record what actually happened: what I believed at the time, what I tried, what happened next, and what I would do differently now.

This is the first note in that record.