FinMind

Inspiration

Managing my personal finances has always been a challenge. I've tried a ton of apps, excel sheets and even hand-written notebooks to find myself disengaged within the first month of managing my finances. That's when my Duolingo streak widget came into sight, my streak of 500+ meant that I had at least been doing 1 language exercise a day for 500 days straight. A simple feature that kept me engaged with learning for almost two years, so what if I people could manage their finances in a similar manner? That's how FinMind came to life: a simple yet engaging platform where people could register their spendings and earnings, with visual, traceable rewards to keep then disciplined.

What it does

To put it simply, log in your personal finances daily to keep your streak alive, while reaching the app-set targets to gain more experience. By registering your earnings and spendings, you'll be able to get a daily financial score, calculated in this first version with a simple algorithm:

1. Daily Budget Calculation First, it calculates your Daily Budget, which is simply your total Monthly Expenses divided by 30 days (defaulting to $100/day if no profile is set up).

2. Base Score The calculation starts with a neutral base score of 50 points.

3. Spending Habits (Up to +20 points or -20 points) Your daily spending is compared against your daily budget as a ratio (spent / dailyBudget):

< 80% of budget spent: +20 points < 100% of budget spent: +10 points

120% of budget spent: -10 points 150% of budget spent: -20 points

1. Savings Habits (Up to +20 points) The app rewards you for putting money aside:

Any savings at all (> $0): +15 points Significant savings (> 20% of your daily budget): +5 bonus points (Total +20)

1. Income Boost (+5 points) If you log any income for the day (> $0), you receive a small boost of +5 points.

2. Mental Wellness / Stress Penalty (-5 points per worry) For every financial "worry" or stressor you select during the check-in (e.g., Debt, Bills, Groceries), it deducts 5 points.

3. Final Output Finally, the algorithm takes the sum of these adjustments and clamps it so the score never drops below 0, and never exceeds 100.

How we built it

The app/platform was built with AI-assisted programming, by leveraging agentic coding tools such as Google's Antigravity (newly released) with Gemini 3.1 Pro model. I provided a basic boilerplate of how I wanted the app's workflow and the languages and frameworks. The app as of now is designed as a progressive web app with mobile-first development in mind. The decision was made as building a web app means simpler hosting and accessibility across multiple devices and only requires an internet connection. For the frontend and UI React was used over a Vite framework, while the backend is managed by Firebase (authentication and database).

Challenges we ran into

The main challenges came when understanding what would work and not for the end user. I didn't want to build a cluttered app that the user was only going to use once. The motto was to keep it as simple as possible. The engagement should be happening in the first minute, and the whole workflow should be executed within 5 minutes at most.

Accomplishments and learnings

On a developer perspective, I'm proud of being able to leverage AI tools to build the application. These days it is easy to only prompt an idea and let the AI do the magic, forgetting the behind the scenes of all the development going on. By leveraging AI tools and human ideation and validation, I think it led to a more complete outcome to what only the AI is able to generate.

On a product development perspective, having testers tell me that the app feels simple yet engaging to keep it in their daily lives was a huge accomplishments. Overall while developing products I think the hardest part is to really understand what the end user needs and having all those features ready in the product.

What's next for FinMind

I'd love for FinMind to be tested across different demographics, to see what's the possibilities of improvement. As of today the testing only occurred within my social circle, but expanding to people with different backgrounds, ages, and spending patterns could really help us see which are the areas of improvement.

Built With

• css3
• firebase
• javascript
• react
• tailwind