About the Project
Inspiration
I wanted a tailored wellness regime for myself and couldn’t find it all in one place — hence the idea for CycleSyncAI. Women’s fitness and nutrition needs fluctuate across their menstrual cycle, but most mainstream wellness apps ignore this. I was inspired to build an app that would fill this gap by providing evidence-informed, personalized recommendations for each phase.
What it does
CycleSyncAI delivers personalized, cycle-aware diet and fitness plans — along with a ready-to-use grocery list — to help women align their wellness routines with their body’s natural rhythms.
It uses menstrual cycle data from Apple HealthKit, combined with user details (age, height, weight, dietary restrictions, medical conditions, fitness goals, and preferences), to generate expert-informed recommendations via Perplexity’s Sonar Pro API.
Specifically, the app provides:
- Personalized diet suggestions
- A daily grocery list aligned with the diet
- Personalized workout plans
- Motivational and supportive messages
How I built it
I built CycleSyncAI as a native iOS app using:
- Swift and Xcode for the frontend and backend.
- Apple HealthKit to access menstrual cycle phase data.
- Perplexity Sonar Pro API to power the core AI functionality — generating diet plans, workout suggestions, motivational feedback, and grocery lists.
- UIKit for interface design, including gradient backgrounds, rounded cards, and smooth animations.
- Local data handling (user profile inputs) to customize the LLM prompts.
- GitHub for version control and project management.
I combined my data science skills (for prompt crafting and shaping user inputs for the LLM) with newly learned iOS development techniques to create a complete, user-friendly app.
Challenges I ran into
Frontend development was completely new to me as a data scientist — my main coding experience is in Python, not Swift.
I faced many challenges learning and polishing the iOS UI, including layout constraints, animations, Xcode configuration, and ensuring a smooth experience across multiple screens.
Integrating Apple HealthKit was another hurdle, especially handling menstrual cycle data securely and testing permissions carefully.
Finally, crafting structured, meaningful prompts for the Perplexity API required iterative experimentation to get useful, relevant outputs.
Accomplishments that I'm proud of
I’m proud of combining a wellness-driven idea with cutting-edge technology (an LLM-powered backend) to create a working iOS app that’s both functional and meaningful.
I’m especially proud of successfully integrating Apple HealthKit — a platform I’d never worked with before — and delivering an app that personalizes suggestions in a way most mainstream apps overlook.
What I learned
I learned how to work outside my comfort zone, diving into iOS app development and Swift.
I gained hands-on experience with HealthKit integration, UIKit design, Xcode project handling, and GitHub version control.
Additionally, I deepened my understanding of prompt engineering and integrating LLM-based APIs into a real-world application, learning how to shape user data into effective queries and process the returned results meaningfully.
What's next for CycleSyncAI
The next steps are:
- Speeding up the response time by exploring partial or streamed LLM outputs to reduce waiting time for the user.
- Expand to generate a personalized diet plan for the entire week (instead of just for the day), including a consolidated grocery list for the upcoming week — making it more convenient and practical for users to plan ahead, shop once, and stay on track.
- Expanding to Android to make the app accessible to more users.
- Incorporating richer data sources, such as fitness tracker or wearable inputs, to further personalize recommendations.
- Refining the UX to make the app more interactive and engaging, possibly adding progress tracking, reminders, or community features.
Potential Impact
This app empowers women to better align their nutrition and workouts with their body’s natural hormonal cycles, potentially improving energy levels, fitness outcomes, and overall well-being.
By offering cycle-aware health suggestions, it promotes mindful, personalized self-care, supports healthier body image, reduces decision fatigue through ready-to-use grocery lists, and encourages positive motivation to eat well and stay active — including kind, supportive messages delivered directly in the app.
Log in or sign up for Devpost to join the conversation.