For anyone with too many submissions
Hundreds of submissions. A shortlist you can defend.
Applications, essays, pitches, grant proposals, creative work. AI judges compare entries head to head, the way a panel would, and produce a ranking with uncertainty, not a score out of ten.
Join the waitlistHow it works
Bring the pile and the criteria
A folder of documents and a short description of what "good" means to you.
Judges compare pairs
Each comparison is run in both orders to cancel position bias, and spent on the close calls where it matters most.
Get a ranking you can inspect
A ranked list with intervals, the reasoning behind each comparison and checks on how consistent the judges were.
Why pairwise, and why error bars
People and AI models are both better at saying which of two things is better than at giving a reliable absolute score. Comparing pairs and fitting a ranking over all of them is the method behind chess ratings and many AI leaderboards.
A ranking without uncertainty invites over-reading small gaps, so every ranking shows which entries are clearly separated and which are effectively tied. We are testing it on public writing first.
Join the waitlist
We're building this now. Tell us what you'd simulate and we'll reach out as early access opens. The most interesting use cases get in first.