Eran Shayshon

I spent a year teaching AI to read a mind map like a human argument — here's what I built

by

Most mind-map software is a drawing tool dressed up as a thinking tool. You arrange nodes, connect ideas, cluster themes — and then the canvas just sits there, mute. The structure you built already encodes an argument. Nothing reads it back to you.

That gap bothered me enough that I spent the last year building Sensemaker — a mind-map app where the AI reads both your text and your spatial layout: what you clustered together, what you connected, what you pushed to the margin. It then writes you back a 200–600 word narrative that names the themes, surfaces the tensions, and ends with the conclusion your map is implicitly building toward.

A lot of rewrites later, I'm finally happy enough with it to tell people.

If you're someone who uses mind maps for strategy docs, research stacks, or prep for talks — I'd genuinely love to know what's broken. Free tier, no card required, multilingial with RTL support.

Launching May 28th on Product Hunt: https://www.producthunt.com/prod...

12 views

Add a comment

Replies

Be the first to comment