Momentum AI

• 

  AI Chat Interface

• 

  Landing Page

• 

  Login

What inspired me

The inspiration for Momentum AI came from my own experience as an introvert who found gyms intimidating. I wanted a fitness guide I could talk to anytime without judgment—something that understood my goals and busy schedule. I realized many people likely felt the same, so I built the solution I couldn't find: a personal, always-on AI companion that makes fitness accessible and stress-free.

How I built it

Momentum AI is a full-stack application built with a focus on a great user experience and powerful AI.

• Frontend: React & TypeScript with Tailwind CSS, hosted on Netlify.
• Backend: FastAPI (Python), hosted on Replit.
• Database & Auth: Supabase for all user data, authentication, and chat memory.
• The AI Brain: Anthropic's Claude 3.5 Sonnet, using tool-calling to intelligently generate conversational advice or full workout plans.

Challenges I faced

This hackathon was a rollercoaster of real-world debugging under pressure. In the final 72 hours, I tackled a series of critical bugs, including a faulty premium-tier feature causing 403 errors, a 500 Internal Server Error on the backend after a code cleanup, and a frontend race condition that broke the user onboarding flow. Each challenge required digging deep into the full stack—from the database to the frontend state—and taught me an incredible amount about building and stabilizing a real product.

What I learned

Beyond the technical skills, this project taught me the critical importance of prioritization under pressure. I learned that a stable, polished, and bug-free core experience is far more impressive than a dozen half-finished features. This hackathon was less about just writing code, and more about the real-world journey of building, debugging, and delivering a product you believe in.

Built With

• 3.5
• anthropic
• apis:
• claude
• fastapi
• languages:-typescript
• netlify
• python-frameworks:-react
• replit
• sonnet
• tailwind-css-platforms-&-services:-supabase