I built a resume tool. Then I used it to get into grad school.

I built FastCV to solve a problem I kept watching other people struggle with: a strong resume that says nothing about the specific job it’s applying for.

Most people write one resume and send it everywhere. FastCV rewrites it for each job description, matching language and surfacing the right experience without inventing anything.

I never planned to be the test case.

The first test was job applications

Before any of the grad school part, I ran it on myself the ordinary way. Ten job applications, each one with the resume tailored to that posting rather than the same file sent ten times. Four of them came back as interviews, and all of that happened inside ten days.

That was the first evidence the thing worked on someone other than a hypothetical user. It was also the first time I could see which part of the output was doing the work, because I was reading the same rewrite the recruiter was.

Being your only user is a real stage

For a while I was the only person using it, and I used to treat that as a weakness. It is not. It is the stage where feedback is fastest, because there is no gap between noticing a problem and being the person who can fix it.

The more I used it, the more it changed. Early versions rewrote too aggressively and produced something that read like a job posting with my name at the top. Recruiters can smell that. So the rewriting got more conservative, closer to my own phrasing, and the tool moved toward surfacing the experience that already matched instead of dressing up experience that did not.

Every one of those changes came from using it on an application I actually cared about. I would not have found them from a survey.

Then the application that mattered

Earlier this year I applied for the MS in AI Engineering at Quantic School of Business and Technology. I had a BSc in Information Technology and an MSc in Computer Science, a handful of projects, and no clear idea how to package any of it for a scholarship committee.

So I did the same thing again. I fed it the program description, tailored my resume to what the application actually asked for, and rewrote my story around the work instead of the credentials.

It worked. I got in with a full scholarship, Class of 2027.

What I took from it

The lesson wasn’t really about the tool. It was that most rejection isn’t a skills problem, it’s a presentation problem. I had the experience the whole time. I just hadn’t written it down in a way that matched what the reader was looking for.

The other lesson is about how the tool got good. Not from a plan, from repeated use by someone who felt the result. If you are building something and you are the only person on it right now, that is not a signal to stop. It is the cheapest research you will ever get.