About this project
How it got here
The problem
Every decision tool asks you to rate how much each criterion matters. But if you already knew your own priorities, you wouldn't be struggling with the decision.
Stage 1
The blank form
Options, criteria, weights — all manual. The recommendation is arithmetic on your own assumptions, fed back at you.
"Sliders assume you already know your priorities. That's the thing we're trying to solve."
Removed the blank form
Stage 2
AI extracts the starting point
Describe the decision in plain language. AI extracts options and criteria. You still edit everything — but you're editing a starting point, not building from nothing.
"Sliders still ask you to consciously rank priorities you might not consciously know."
Weights deduced, not declared
Stage 3
Scenarios replace sliders
Two concrete outcomes, one hidden tradeoff. Pick the one you'd rather live with. Six questions reveal your priority hierarchy without ever asking you to name it. Conjoint analysis.
"Choices can conflict — prefer A over B, B over C, but C over A. Simple tallying breaks down."
Circular preferences handled
Stage 4
AHP resolves contradictions
AHP detects inconsistent choices and fires one targeted follow-up. Weights emerge from your behaviour, not your self-assessment.
"Scenarios used named characters — Priya, Tariq. You were watching someone else's life, not living the tradeoff."
You're in it, not watching it
Stage 5
Second person wording
You made this choice. This is what your situation looks like now. The tradeoffs land differently. Sliders remain — drag one and watch the result change.