Money Bags

Problem it solved

Résumé scorers average their checks, so one fatal flaw hides behind two good scores.

A scorer with three lenses (parser, AI screener, human read) where the lowest lens is the score.

AI workflow

• Each lens scored by its own prompt, with reasons tied to a line in the résumé
• Scoring rules locked as ADRs before the build
• Type and layout set in Figma, matched in code