WillApp

Define your Wills. Prove it with AI. · Academic Goal & Study Engine

Production Beta Flutter Cloud Firestore Gemini Multimodal AI

Project Summary & Problem Statement

University students struggle to maintain self-directed study schedules when preparing for rigorous examinations. WillApp bridges intent and execution through automated syllabus parsing, milestone breakdowns, and tamper-resistant AI verification.

Bruno's Role & Engineering Contributions

Role: Co-founder & Lead Backend Engineer within a 5-person engineering & product team.
Study Engine Pipeline: Architected the 4-step wizard that ingests course syllabi or exam deadlines and synthesizes structured study roadmaps into Cloud Firestore.
Multimodal Verification: Built serverless Google Cloud Functions that evaluate student-submitted study proof (handwritten notes, problem sets) against task specifications via Gemini Multimodal API.
Gamification & Retention: Implemented streak mechanics, XP tiering, and server-side anti-cheat heuristics.

Architecture & Evidence

Client Layer:      Flutter SDK (iOS & Android) with Liquid Glassmorphism UI
Cloud Functions:   Node.js / TypeScript microservices with Firebase Admin SDK
Database:          Cloud Firestore with strict role-based security rules
AI Pipeline:       Gemini 1.5 / 2.0 Multimodal Visual Inspection Engine
Authentication:    Firebase Auth (Google Sign-In, Apple Sign-In)

Known Limitations & Privacy Controls

Verification Scope: Multimodal AI inspection is heuristic; appeals for edge cases are handled through in-app review channels.
Privacy First: Uploaded validation photos are stored under authenticated tenant storage rules and are never shared publicly or used for model training without explicit consent.