From 0 to 100 Users: How I Built an AI Humanizer as a Student

Most "humanizers" break your text by rewriting it. I built a tool that bypasses AI detection while keeping the content 100% identical. Concealy (https://concealy.org) makes AI text undetectable without changing a single word. Here's how I found my first 100 users via Telegram and friends. 👇

The problem: Rewriting breaks everything
As a student, I saw everyone panicking about AI detection. The existing solutions were terrible. They worked by "paraphrasing," which meant they swapped words, ruined sentence structure, and often changed the meaning entirely.

If you had a direct quote or a snippet of code, those tools would mangle it. You’d get a green checkmark from the detector, but your essay would read like garbage.

I didn't want a re-writer. I wanted a bypasser.

The solution: Invisible changes
I built Concealy with one rule: Input must equal Output.

I developed a method that tweaks the text at a structural level, not a linguistic one.

No synonyms swapped.
No grammar altered.
No tone shifts.

Visually, nothing changes. But to detectors like GPTZero or Turnitin, it reads as human. This is huge for coding or technical writing where changing a specific variable name or term breaks the whole project.

Getting users: Friends and Telegram
I didn't run ads. I didn't have an audience. I started with who I knew.

I shared the link with my local friends and dropped it into my University Telegram group. These were people already stressed about deadlines.

I didn't pitch it as a product launch. I pitched it as a favor.

Asking for feedback, not sales
This was key. If I had gone into the group saying "Buy my SaaS," they would have ignored me.

Instead, I said:
"I built this tool to stop AI flags without ruining the text. It doesn't change a single word. Can you guys test it and tell me if it works for you?"

People love giving opinions more than they love spending money. They tried it because they wanted to prove me wrong or help me out. Once they saw it actually worked—and that their essays remained intact—they kept using it.

I also threw it on Product Hunt to get some eyes outside my local circle, but the core 100 users came from that initial push for feedback.

Making money
I'm not doing this just for fun. My motivation is to build a profitable business.

I currently have 3 pricing plans set up. The free tier gets them in the door, but the paid plans are where the sustainability is.

Basic: For occasional users.
Pro: For heavy users/students.

What's next?
I'm at 100 active users, but I want to scale. My goal now is to turn those free feedback users into paid subscribers and expand beyond just my university circle.

If you are writing AI content and can't afford to have your words changed, check out Concealy and let me know what you think.