Why we are building it
The mark is the easy part. Knowing why is the part nobody has time for.
Essay written
An hour of thinking, handed in and out of your hands.
Marked, eventually
A number at the top of the page, a fortnight after you wrote it.
Same gap, next essay
Whatever cost the marks is still there, unnamed and unfixed.
A 9708 essay is not marked on whether the answer is right. It is marked on knowledge, application, analysis and evaluation — and a script can be full of correct economics yet still lose marks because the evaluation never arrives.
Teachers know this better than anyone, but detailed feedback takes hours. So feedback comes back thin, late, or both, and the next essay repeats the same mistake.
Not a tutor. Not a replacement for a teacher. One thing, done carefully.
Marginal gives an essay back with an assessment-objective breakdown: what earned marks, what did not, and what a stronger answer would have done differently.
Who is building it
Two co-founders, and no one else yet.
Both profiles are public — worth a look before you give an early product your email address.
What Marginal is not
Two things worth saying plainly.
It is independent.
Marginal is not affiliated with, endorsed by, or connected to Cambridge Assessment International Education. The syllabus code describes what the tool marks, and nothing more.
It can be wrong.
Marking is done by AI, and AI can make mistakes — on marks, feedback, and reading text from a photographed script. Every mark is a second opinion, not a verdict.
The longer version is in the terms and privacy policy.
Want it when it opens?
One email when the first cohort opens. Nothing else.