Cardi-Oh

A smart fitness app that builds custom workouts based on your body and gear, with YouTube playback, AI compliments from a “hot” voice, and a pet that lives only if you stay consistent.

Comment

Inspiration

Like many students, we started working out but couldn’t stay consistent. Either we didn’t know what workouts to do, or we lost motivation quickly because results took too long. So we thought — what if we had something that personalized a plan for us and kept us emotionally invested? That’s how Cardi-Oh was born.

What it does

Cardi-Oh is a personalized fitness app that recommends workouts based on your fitness level, body type, goals, and available equipment. To help users stay consistent, we gamify the experience with a virtual pet or plant that thrives only when you complete your workouts. You can paste a YouTube video to watch while exercising, and at the end, you're rewarded with AI-generated compliments from a voice that feels like your gym crush hyping you up.

How we built it

Right now, we’ve built a mock authentication system (login/signup) and a main dashboard using:

  • Frontend: Next.js with Tailwind CSS
  • Backend: Supabase (PostgreSQL) + Next.js API routes
  • Realtime-ready: Setup for future pet state tracking with Supabase Realtime
  • Deployment: Vercel for frontend, Supabase for backend

We designed the architecture with scalability in mind so features like the AI plan generator, pet state system, and YouTube workout flow can be added quickly.

Challenges we ran into

  • Balancing ambition and scope — we had a lot of exciting features, but time limited how much we could ship
  • Learning Supabase’s auth and database system on the fly
  • Making sure the dashboard experience was smooth even before all core features were in

Accomplishments that we're proud of

  • Set up a clean, scalable project architecture
  • Built mock auth and a functional dashboard from scratch
  • Designed a full user flow and mapped out the logic for gamification and personalization
  • Laid a solid foundation for real-time updates, AI workout generation, and YouTube integration

What we learned

  • Supabase is incredibly powerful for fast prototyping when used right
  • Next.js + Supabase is a killer combo for full-stack apps
  • Sometimes, it’s better to nail the core experience and leave the extras for after the hackathon

What's next for Cardi-Oh

  • Build out the fitness input form and workout plan generator
  • Add YouTube timer integration for workout playback
  • Implement pet/plant gamification with real-time state updates
  • Use AI-generated audio compliments for fun feedback
  • Eventually: ML-based workout personalization and social features (friend streaks, challenges)

Built With

Share this project:

Updates