Building with users, not for them

I built the first version of EduVision in three weeks and I was proud of it. It read textbooks aloud for blind students. I thought I understood the problem. I was wrong.

The first time an actual blind student tried it, he fumbled for a long time. It turned out that what I considered “convenient” — neat buttons laid out on a screen — was invisible to him. He needed to hear where he was, not see it. I sat still for the rest of that afternoon, ashamed that I had designed for a person I imagined instead of the person sitting in front of me.

After that I changed how I work. Before writing a single line of code, I go and ask. I record the conversations and listen back for the moments people hesitate — because hesitation is where the real problem hides. The product looks worse on paper now, and works better in real hands.

The lesson sounds obvious, but I had to break something once before I truly believed it: you don't build for users, you build with them. The person who understands a problem best is always the one living with it, not the one writing the code.