AiLedge

Record and spend smarter with AiLedge

Comment

Inspiration

Managing personal finances and tracking daily expenses like groceries can often feel tedious and time-consuming. We wanted to build a financial tool that goes beyond traditional budgeting apps by leveraging the power of Artificial Intelligence. The inspiration for AiLedge was to eliminate the friction of manual data entry and provide users with a smart, offline-first assistant that effortlessly tracks their spending, analyzes their financial health, and offers personalized insights.

What it does

AiLedge is an AI-Powered Personal Financial Planner & Grocery Spending Recorder. It serves as a comprehensive hub for all things finance, offering:

  • AI Receipt Scanning: Users can take a photo of their receipts, and the app leverages Gemini AI to automatically extract itemized expenses and categorize them.
  • Smart Financial Chat: A built-in AI assistant helps users answer financial queries and provides tailored advice based on their spending habits.
  • Comprehensive Dashboard: Interactive charts and analytics to visualize spending vs. budgeting.
  • Secure Data & Passcode Lock: Built with privacy in mind, all financial data is stored securely and efficiently on the device with a biometric/passcode lock.
  • Budget & Recurring Transactions: Tools for managing monthly budgets, tracking groceries, and handling recurring subscriptions.

How we built it

  • Framework: We built the cross-platform mobile application using Flutter and Dart.
  • State Management: Handled via the Provider package for scalable and reactive UI updates.
  • Artificial Intelligence: Integrated Google's Gemini LLMs via the google_generative_ai package to power the Receipt OCR parsing and the intelligent chat assistant.
  • Local Storage: Used Hive for blazing-fast, NoSQL local data persistence, ensuring an offline-first experience.
  • UI/UX: We incorporated modern design libraries like fl_chart for dynamic data visualizations, and animate_do alongside flutter_staggered_animations for smooth, responsive user interactions.

Challenges we ran into

  • AI Prompt Engineering: Tuning the Gemini model to consistently and accurately parse messy, differently formatted store receipts into structured transaction objects required heavy iteration.
  • Offline Data Synchronization: Designing a robust schema using Hive to handle complex relational data (Accounts, Transactions, Budgets) while keeping queries fast and syncs seamless.
  • Performance & Processing: Managing state effectively when processing heavy image files for receipt scanning without blocking the main UI thread.

Accomplishments that we're proud of

  • Successfully integrating cutting-edge Generative AI into a local mobile client to completely automate expense tracking.
  • Delivering a premium, dark-themed user interface packed with micro-animations that make financial planning feel engaging rather than like a chore.
  • Keeping the application completely offline-first, ensuring that sensitive user financial data never leaves their device unencrypted.

What we learned

  • Advanced state management techniques for scaling a Flutter app with varying data sources.
  • Practical applications of LLMs (Large Language Models) beyond chatbots, specifically utilizing them as powerful natural language and image processing engines.
  • How to properly manage local lifecycle events and app security locking mechanisms in Flutter.

What's next for AiLedge

  • Cloud Sync & Backup: Implementing end-to-end encrypted cloud backups so users can access their financial data across multiple devices.
  • Bank Integrations: Adding support for Plaid or similar APIs to automatically pull in credit card and bank transactions.
  • Advanced Investment Tracking: Expanding the app's scope from daily expense tracking to long-term portfolio and crypto tracking.
  • Multi-language and Multi-currency Support: Opening up the app to a global user base.

Built With

Share this project:

Updates