Every tool I tried would jump straight to screens. But that's not how product design actually works. You don't just design screens. You think through the problem first. The flows, the edge cases, the user journey, where people will get stuck. Then the design comes finally.
Figr does that thinking layer first. It parses your existing product via a chrome extension or takes in screen-records, then works through the problem with you before designing. Surfaces edge cases, maps flows, generates specs, reviews UX. The design comes after the thinking.
It is able to do so because we trained it on over 200k+ real UX patterns and UX principles. Our major focus is on helping in building the right UX by understanding the product.
The difference from Lovable/Bolt/V0: I think those are interface builders. They are good when you know exactly what you want to build but they don't truly help in finding the right solution to the problem. Our aim with Figr is to be more like an AI PM that happens to also design.
Some difficult UX problems we've worked through with it: https://figr.design/gallery
Would love feedback, especially from folks who've hit the same wall with other AI builder/design tools.