Balance

Inspiration

The idea for Balance was born out of a personal struggle: trying to juggle productivity, fitness, family time, and social life without burning out. I wanted an app that didn’t just remind me of tasks, but actually understood my context — where I was, what I had coming up, and what mattered to me long-term. I tried habit apps, but the data input was demanding because I needed to tell it what I did, where I was, and when, instead of just getting suggestions of what would help me achieve my goal (where, when & why) The spark came from a question I asked myself:

“What if an app could act like a guide, nudging me into better habits at the right place and the right time?”

What it does

Balance is an AI-powered well-being app that helps you build and maintain healthier routines. It looks at your goals, schedule, and location to suggest the right activities at the right time — whether that’s hitting the gym, focusing on deep work, spending time with family, or even preparing for social commitments. It also brings in social accountability through peer feedback and check-ins, so you stay consistent. In short: it’s your personal guide to productivity, health, and work–life balance.

How we built it

Frontend: React + React Native, with clean UI components to keep it simple and modern. Backend: NestJS with MySQL (with an eye on Postgres later). AI Integration: OpenAI function calls to generate weekly activity schedules tailored to goals and locations. Extra Magic: Background location tracking (scheduled triggers, even when the app is closed). Complete tasks in the background if you go to a location an activity was scheduled for I leaned into rapid iteration — building core logic first, then layering design, features, and accountability tools on top.

Challenges we ran into

1. App Store Approval: Google Play was smooth, Apple is stricter and slower — patience is key.
2. Tech Hurdles: Managing background services in React Native after ejecting from Expo. Finally plugged in via custom scripts to modify Android manifests and iOS storyboards Ran into a classic N+1 problem in the back and had to modify data queries
3. Community Building: Without Reddit karma or Discord groups, finding places to share the app has been tricky.
4. Balance in Balance: Ironically, working on a “balance” app meant I had to constantly check myself not to burn out in the process while pushing the product.
5. I signed up for RevenueCat recently and only saw the competition on the 25th of September so I had to rush the integration for in-app purchases
6. Getting characters consistent using vio3 was a bit challenging
7. OpenAI had a tendency not to be on course, so I had to do proper research on prompts. Implementing the stream was a bit hard for me to show how the AI is thinking on the app
8. Built the core functionality, but it still requires design polishing
9. I come from a very backend world, and getting the design right was always a struggle, but I have been trying to learn

Accomplishments that we're proud of

Approved on Google Play, with iOS review pending Built a full-stack app with React Native frontend and NestJS + MySQL backend Integrated AI-powered activity scheduling tailored to user goals and context Implemented background location tracking that works even when the app is closed Migrated from AsyncStorage to SQLite for better local storage scalability Designed and rolled out a modern, user-friendly UI/UX redesign Created a social accountability feature (“Seesaw Sessions”) for peer reviews and motivation Developed a unique AI activity engine that generates personalized routines Getting the right balance of background location tracking while balancing battery life, and knowing to throttle location fetching when walking or stationary, and tracking when you are on the move, so that you collect location more regularly

What we learned

Building Balance wasn’t just coding — it was an education in psychology, design, and systems thinking. Some key takeaways: Habits are social. People stick to goals when they’re accountable to someone else. Context is king. A reminder to “go to the gym” is meaningless if I’m stuck at work. AI can be a coach. With the right prompts and structure, AI helps generate personalized activity suggestions that feel natural. I also explored how consistency compounds over time, similar to exponential growth

What's next for Balance

1. Expand from Android to iOS launch once Apple approval is complete
2. Roll out premium features like Seesaw Sessions (peer accountability system)
3. Add smarter notifications tied to context (e.g., location, upcoming events, traffic)
4. Refine the AI activity engine with more personalized recommendations and better scheduling logic
5. Build a referral system to boost organic growth through friends and accountability partners
6. Continue bug fixes and rapid iterations based on real-world user feedback
7. Explore integrations with calendars, wearables, and health apps for richer insights and become a center of human behavior
8. As I was integrating, I started thinking of things that can make it sticky and insight into where and how you spend your time to make you realise if you are living healthily or not

Built With

• mysql
• nestjs
• react-native