MoneyMaxxing

Inspiration

Gen Z is constantly bombarded with spending temptations — from TikTok trends to late-night fast food cravings. Most budgeting apps are either boring, overly complex, or made for boomers. We wanted to create something that actually vibes with Gen Z: a smart, fun, and intuitive budgeting app that speaks our language, helps us save, and still lets us enjoy life without guilt.

What it does

MoneyMaxxing is a Gen Z-first budgeting app that transforms how young users interact with their money. It combines clean design, intelligent automation, and AI-powered insights to make managing personal finance easy, intuitive, and even fun.

Key Features:

• 📊 Interactive Budgeting Dashboard
  See your finances at a glance with beautiful, dynamic graphs that break down spending, income, and category-wise expenses. Everything updates in real time as you add or edit entries.

• ➕ Effortless Income & Expense Tracking
  Users can quickly log purchases or income with just a few taps. Each entry is editable and organized, with smart categorization to help users stay on top of where their money goes.

• 💸 DPR (Daily Purchase Rate)
  A core financial guidance feature, DPR uses TensorFlow Lite(LiteRT) to analyze your past spending habits and calculate how much you can safely spend per day. It's like having a smart spending limit that adapts to your lifestyle.

• 🤖 SmartSpend – AI-Powered Purchase Advice
  This is your personal finance advisor in your pocket. Enter what you want to buy, how much it costs, and how badly you want it. SmartSpend analyzes your remaining budget, how the purchase impacts your daily rate, and uses Gemini API to give a 1–10 recommendation rating with a short, personalized explanation.

How we built it

• Frontend & UI:
  Built entirely using Flutter, allowing us to create a smooth, cross-platform mobile experience with a modern and responsive interface tailored for Gen Z users.

• State Management & Local Storage:
  We used SQLite for on-device data storage, managing all budget entries, income, and expenditure data through a custom DatabaseHelper. This gave us full control over how transactions are tracked and updated, even offline.

• Machine Learning with TensorFlow Lite:
  We integrated TensorFlow Lite (LiteRT) to calculate the Daily Purchase Rate (DPR) based on the user's historical transaction data. The model runs entirely on-device to ensure performance, privacy, and a seamless experience.

• AI-Powered Recommendations with Gemini API:
  Our SmartSpend feature is powered by the Gemini API, which receives structured inputs (like item cost, desire level, and financial status) and returns a rating (1–10) with a concise explanation of whether the purchase is advisable. The prompt structure and response parsing were carefully designed for consistency and clarity.

• User Experience:
  We focused heavily on building an intuitive and engaging UI/UX — including color-coded visuals, real-time DPR updates, interactive charts, and a clean interface that simplifies complex financial information.

Challenges we ran into

One of the biggest challenges was getting TensorFlow Lite to work properly within a Flutter environment. Deploying the DPR model on-device using LiteRT required digging through a lot of documentation, managing platform compatibility, and optimizing the model to run smoothly without slowing down the app.

Another major hurdle was in the early tech stack decision. I initially tried building the app using React Native with Expo, thinking it would be quicker. But I ran into multiple compatibility issues, especially when integrating native features like local storage, ML model inference, and graph rendering. Things got messy fast.

I always envisioned this as an app-first experience — a website just didn’t feel personal or seamless enough for something like MoneyMaxxing. Budgeting is deeply personal, and I wanted users to feel like the app lives with them — on their phones, with their data, in real time.

So I made the tough call to switch to Flutter mid-build. It was a huge reset, but Flutter gave me the control, performance, and design consistency I needed. In the end, it was 100% worth it.

Accomplishments that we're proud of

• 🧠 Successfully integrated TensorFlow Lite to compute the Daily Purchase Rate (DPR) entirely on-device, with smooth performance and real-time updates.
• 🤖 Built a fully functional SmartSpend feature using the Gemini API — delivering personalized, AI-powered purchase recommendations with contextual reasoning and dynamic scoring.
• 📊 Implemented 90% of the envisioned graphs and analytics, including budget breakdowns, spending trends, and DPR impact visualizations — all interactive and visually clear.
• 🔁 Transitioned tech stacks mid-hackathon from React+Expo to Flutter, and still delivered a polished, mobile-first app experience.
• 🎯 Brought the core vision to life — an intuitive, Gen Z-focused budgeting app that actually helps people think before they swipe.

What we learned

• 🛠️ Flutter Mastery (in 36 hours!)
  Coming into the hackathon, I had only scratched the surface of Flutter. After switching from React+Expo mid-project, I had to ramp up fast — and by the end, I was confidently building custom UI components, managing app state, handling async storage with SQLite, and integrating native plugins. Flutter turned out to be the perfect tool for a polished, performant, mobile-first app.

• 🤖 Integrating GenAI into a real-world mobile app
  I learned how to structure prompts, send dynamic context to the Gemini API, and parse its responses in a meaningful way. I also had to make the AI's reasoning feel helpful and human — not robotic.

• 📈 ML On-Device with TensorFlow Lite
  This was my first time using TensorFlow Lite, and it pushed me to understand how to optimize models for mobile, reduce inference lag, and structure input/output in a Flutter environment using LiteRT.

• 🎨 Importance of Design and User Flow
  I realized that no matter how smart your backend is, it doesn’t matter if users can’t understand or enjoy using the app. I spent time making sure MoneyMaxxing felt clean, modern, and easy to interact with — especially for a Gen Z audience.

• 🧩 Scoping & Pivoting Wisely
  I learned to pivot when needed (from React to Flutter), cut non-essential features, and double down on what mattered most: core functionality, good design, and a meaningful user experience.

What's next for MoneyMaxxing

• 🔗 Real-time bank integration
  We'll integrate APIs like Plaid to pull in live transaction data, so users won’t have to manually log their expenses — making budgeting even more effortless.

• 🎮 Gamified budgeting features
  We're planning features like “Spending Streaks,” achievement badges, and financial challenges to keep users motivated and make smart spending a habit.

• 🧠 Smarter SmartSpend
  We'll expand SmartSpend to support more natural questions like:

  “Can I buy this if I skip coffee for a week?”
  And allow it to learn from user feedback to improve recommendations over time.

• 📦 Cloud sync & cross-device support
  Adding secure backup and syncing across devices so users can budget from anywhere.

• Possibly become the next viral Gen Z fintech app? 👀

Made with 💻, ☕, and ❤️ at HackPrinceton S'25 by Sherwin Vishesh.

Built With

• dart
• flutter
• gemini-api
• sqlite
• tensorflow-lite