As Databricks scaled fast, sustaining engineering quality depended on people who actively raise the bar - for code, for design, for hiring, and for what the team considers acceptable - rather than merely clearing it themselves. Raising the bar means being a multiplier: leaving code, systems, and colleagues better than you found them; holding a high standard in review and design without becoming a gatekeeper; mentoring so others reach the bar rather than depending on you; and being honest in hiring and calibration even when it's easier to pass someone through. The value is watching for whether your effect on the org is additive (you did great work) or multiplicative (the team around you got better because of how you work). In behavioral rounds, interviewers look for evidence that you elevate the people and standards around you. The strong signal is a story where you raised a standard that stuck - a review practice, a design bar, a mentee who leveled up, a hiring call you held firm on - and where the improvement outlived your direct involvement. The anti-signals are pure individual-heroics stories with no one else in them, 'high standards' that show up as gatekeeping or nitpicking that demoralizes rather than elevates, and stories where you personally did excellent work but left no durable lift in the team.