Hey everyone 👋 Maker here. I'm an author, not a developer by trade, but I kept running into the same problem with AI answers; they all sound equally confident, whether the claim is rock solid or a total guess. So I built Sift to fix that. Ask it anything, and it breaks its own answer into individual claims, grades each one (solid ground / worth a second look / genuinely unsure), and shows you the reasoning behind every grade. The part I care about most is the Track Record tab. You can rate whether past claims actually held up, so over time it builds a real, checkable history of how reliable it's been. No AI I've used does this. Built it end to end myself over the past few weeks. Would love to know: what's the first thing you asked it, and did the confidence grading match what you already knew? Genuinely curious where it gets it right and where it doesn't.

Sift

An AI that grades its own confidence, claim by claim.

An AI that grades its own confidence, claim by claim.