We fed 50 contracts into an AI. Here's what it found in 11 minutes.

I had a folder. Fifty files. Contracts, a few NDAs, some invoices, and a couple of things I wasn’t even sure what they were. Sorting it by hand would have been an afternoon I didn’t have.

So I did the thing you’d do if you had the right tool: I described what I wanted in a sentence and dropped the folder in.

“Keep only contracts with a non-compete clause”

That was the instruction. Plain English. No regex, no tagging system, no spreadsheet column for “has non-compete (Y/N)” that I’d have to fill in myself.

Eleven minutes later — most of it just waiting — I had a scored list. Every file got a match score and a one-line summary of why it did or didn’t fit. The ones with a real non-compete clause floated to the top. The invoices and the mystery files sank to the bottom where they belonged.

What surprised me

Not that it worked. That it was boring. No drama, no “AI being weird,” no hallucinated clause that wasn’t there. It just did the tedious thing I’d been avoiding, and it did it the way a careful junior associate would — except it didn’t get tired on file 47.

The part that actually changes how you work is the threshold. You set “keep score ≥ 50” and suddenly you’re not reading fifty files, you’re reviewing the eight that mattered. That’s the whole game with document automation: don’t read everything, read the right things.

Try it on the pile you’ve been avoiding

Everyone has one. The folder. The shared drive. The “I’ll get to it” inbox. The reason you haven’t is that sorting it feels like a full day of low-grade pain.

Point an AI at it instead. Describe what you want kept, set a threshold, export the report. You’ll have the answer before your coffee’s cold.

KIRA.id scores up to 50 documents at once against plain-language criteria, on your own infrastructure. See how it works or reach out and we’ll run it on a sample of your files.