• The Engineering Challenge: The client needed an interactive diagnostic tool with strict verification, single-submission data integrity (preventing retakes), and immediate personalized PDF delivery without relying on costly server-side rendering infrastructure. • Zero-Latency Client State & Logic: Formulated a localized scoring state machine that computes weighted persona vectors in real time. Question transitions execute at 60fps with zero network dependency during the assessment flow. • Backend Architecture & Strict Fraud Prevention (Supabase): Architected relational tables for leads, responses, and scoring metrics. Enforced a cryptographic state-lock flag (is_assessment_completed) evaluated at the authentication barrier — stopping re-entry and duplicate submissions even across cache purges. • Serverless Output Engine: Implemented on-device PDF compilation using custom Dart rendering routines. Vector branding, user persona badges, and verified completion timestamps are rasterized directly in-memory and downloaded client-side in under 400ms.