Your rubric, not ours
The fastest way to lose an admissions team’s trust is to hand them a score that reflects someone else’s idea of a good applicant. Programs differ — a conservatory and a nursing school and a CS department are not reading for the same things — and a one-size model flattens exactly the judgment that makes a program a program.
So ProgramApply doesn’t ship with an opinion about who belongs. It learns yours.
How scoring works
You encode your rubric once: the dimensions your committee already evaluates, the weight each carries, and what strong and weak evidence looks like for each. From then on, every file is measured against that — the same bar, applied evenly, whether it’s the first file of the morning or the four-hundredth of the week.
Crucially, the score is never a black box. Each dimension shows:
- the criterion it was measured against,
- the evidence the model found in the file, quoted and linked, and
- where the file fell short, so a reviewer knows where to look.
Consistency is the feature
Humans are wonderful at judgment and terrible at applying it identically across hundreds of files over many tired days. Software is the reverse. The point of rubric scoring isn’t to replace a reviewer’s judgment — it’s to make sure the easy, mechanical part of the read is consistent, so reviewers can spend their judgment where it counts.
Your rubric. Your weights. Your decision. We just make sure it’s applied the same way to file one and file ten thousand.
← All field notes