Pedal

Landing page
Login page
Sign up page
New user survey
Intake tracker page
Recipe suggestion
Full recipe display
Safe Space page
Learn more page
Database admin panel

Inspiration

The inspiration for Pedal comes from Karen Cao’s personal journey with hypothalamic amenorrhea (HA). During recovery, she struggled deeply with her mental health. She often turned to health and fitness apps for support, only to find many were harmful rather than helpful—promoting unhealthy habits and even contributing to the mindset that led to her diagnosis. This project is her way of creating the safe, supportive resource she needed during her journey.

What it does

Pedal helps users recover from hypothalamic amenorrhea (HA) with tools that support both physical and mental health. It features beautiful graphs and uses historical temperature and caloric data to compute the probability of your period returning. Instead of calorie counting, it uses a growing flower petal graphic to track reaching the 2,500-calorie goal, keeping the focus on nourishment rather than numbers.

The app also provides AI-powered meal suggestions, where users can simply upload pictures or describe their food, plus a corkboard and mirror space for adding affirmations and cute stickers to encourage positive self-talk. A Learn More page offers science-backed information and resources on HA, supported by an AI chatbox to answer questions and provide guidance.

How we built it

Frontend (Next.js 14)

React with TypeScript: Type-safe, modern web application
Tailwind CSS: Beautiful, responsive design with custom color palette
Framer Motion: Smooth animations and delightful micro-interactions
Radix UI: Accessible, customizable component library

Backend (FastAPI)

Python FastAPI: High-performance API with automatic documentation
SQLite Database: Lightweight, reliable data storage
Pydantic Schemas: Data validation and serialization
CORS Configuration: Secure cross-origin resource sharing

AI & Machine Learning

Google Gemini API: Advanced natural language processing for meal analysis
TensorFlow/Keras: LSTM models for period recovery prediction
Scikit-learn: Cox regression for statistical health modeling
NumPy/Pandas: Data processing and analysis

Key Integrations

NextAuth.js: Secure authentication with Google OAuth
Nivo Charts: Beautiful, interactive data visualizations
Lucide Icons: Consistent, accessible iconography

Challenges we ran into

We ran into data limitations early on; while variables like bone density are strong predictors for hypothalamic amenorrhea recovery, they aren’t easily accessible through consumer devices, which forced us to rethink our model inputs. Integrating our LSTM prediction model with the live website was another major hurdle, requiring complex coordination between the FastAPI backend and the frontend to make predictions run smoothly in real time. Finally, even small edits to our database schema turned out to be painful, since they required migrating the entire database structure, which slowed down our iteration and testing.

Accomplishments that we're proud of

We’re especially proud of building an AI-powered model that predicts the likelihood of a user’s period returning based on their historical data, turning complex health metrics into personalized, actionable insights. Another accomplishment was rethinking the HA recovery experience to focus on mental well-being rather than numbers. Instead of calorie counts or harsh metrics, we designed a growing flower petal visualization, creating a more encouraging and compassionate way to track progress. This balance of cutting-edge technology with user-centered design is at the heart of what makes Pedal special.

What we learned

Throughout this project, we learned how to manage version control with GitHub, collaborating smoothly while merging different features into one cohesive product. We also discovered how to leverage AI tools to speed up development, from generating starter code to refining complex features. Most importantly, we grew as a team by communicating effectively to brainstorm and refine ideas, and by using empathy and research to truly understand the user’s experience

What's next for Pedal

Looking ahead, we envision Pedal growing into a comprehensive recovery companion for hypothalamic amenorrhea and beyond. One of our top priorities is integrating with healthcare providers, allowing users to securely share progress data with doctors, therapists, or dietitians to enable personalized care. We also plan to add community features, creating a safe space where users can connect with others on similar journeys, share stories, and offer peer support. We hope to fully integrate with wearable devices like Apple Watch and Fitbit so we can make tracking completely seamless for real-time feedback.