The System I Wish I'd Had.
I grew up in Brasília, Brazil. When I decided to pursue a PhD in the United States, I assumed the TOEFL would be easy — just another standardized test. I was wrong.
I took the TOEFL four times. My first attempt was a wake-up call. I studied hard, but I studied randomly — practicing everything equally, panicking about everything equally.
"I walked into my first TOEFL without a clear target score. I just knew I needed 'something good.' That lack of precision cost me months (and $$$)."
After three frustrating attempts, I did what any engineer would do: I stopped studying and started building a system. I reverse-engineered the test. I mapped every task type. I tracked every mistake to its root cause. I measured everything.
That system got me into MIT. Now I've rebuilt it from scratch for the brand-new TOEFL 2026 format — with new task types, new scoring, and new adaptive mechanics. This is that system.