What our admissions committee actually looks for.
It isn't a perfect transcript. It's evidence you'll do the work when nobody is grading it.
Every admissions cycle, applicants ask what actually moves the needle. It's rarely the single highest test score in the pool. We're reading for evidence that an applicant builds things, finishes them, and keeps going after the first version doesn't work.
That shows up in different ways: a robotics project built in a bedroom with parts that didn't quite fit, a dataset scraped and cleaned out of curiosity, a half-finished open-source contribution. We weight demonstrated initiative heavily alongside academic record, because the programs here are project-heavy from the first term, and that instinct to keep building matters more than any single grade.
The strongest applications are specific. Not 'I'm passionate about AI,' but the actual thing that was built, what broke, and what was learned fixing it. If you're applying, spend your essay on that instead of a general statement of interest.