Portfolio of Evidence — 2026
A mobile budgeting application powered by AI receipt scanning and gamification
03 — Theory & Documentation
THEORY & DOCS
Introduction & Problem Statement
Budget management is an essential skill, yet many people struggle to maintain it long-term. With the rise of smartphones, mobile applications have become increasingly popular for managing personal finances. BudgtAI addresses a core flaw identified in existing solutions: user fatigue caused by friction in data entry and a lack of engagement. By combining AI-powered automation with gamification mechanics, BudgtAI transforms a tedious financial chore into a rewarding daily habit.
The goal is not just to track spending — it's to build the habit of financial awareness.
Research & Literature Review
Three leading Android budgeting applications were analysed: Rolly, Goodbudget, and FinWise. Rolly demonstrated the effectiveness of AI-driven expense categorisation and gamification. Goodbudget proved the enduring value of the digital envelope method for disciplined budgeting. FinWise illustrated how advanced forecasting helps users visualise long-term financial health. A critical gap identified across all three: none combined intelligent automation with meaningful user engagement mechanics.
Gaps Identified
No existing solution combined frictionless data entry with genuine financial engagement.
System Design & Architecture
BudgtAI is built on a layered Android architecture: the UI layer (Jetpack Compose), a ViewModel layer managing state via StateFlow, a Repository layer abstracting data sources, and dual storage via Firebase Firestore (remote) and Room (local cache). The AI pipeline connects to Google Gemini via Firebase AI Logic, processing receipt images and generating spending insights on-demand.
Methodology
Development followed an Agile iterative approach across three POE parts: Part 1 established research foundations and design blueprints; Part 2 implemented core features with local storage; Part 3 migrated to cloud storage and integrated AI features. Android Studio was the primary IDE, with GitHub for version control and Jira for sprint tracking.
Research foundations and design blueprints
Core features implemented with local storage
Cloud migration and AI feature integration
Conclusion
BudgtAI represents a modern solution to personal financial management. By focusing on frictionless data entry via AI receipt scanning and sustained engagement through gamification, the design addresses the fatigue identified as the primary failure point in standard budgeting applications. The structural planning documented in this POE demonstrates a clear path from design blueprint to fully functional application.
04 — Technology
TECHNOLOGY STACK
Built with a modern Android stack, powered by AI and real-time cloud sync