HerCycle - Menstrual wellness comapanion

HerCycle is your holistic menstrual wellness companion for menstrual cycle tracking with AI-powered personalized plans for nutrition, movement, and emotional support.

Comment

Inspiration

Menstrual health apps today mostly stop at tracking dates. But periods affect food, energy, emotions, movement, and daily life and most people are left to figure that out on their own.

We wanted to build something that doesn’t just log cycles, but actually thinks with the user and adapts to how their body feels day to day, especially in settings like hostels, tight budgets, or limited access to care.

What it does

HerCycle is a menstrual wellness companion powered by multiple AI agents working together.

  • Predicts upcoming cycle patterns using a machine learning model.
  • Understands symptoms and lifestyle signals.
  • Suggests simple, practical food, movement, and emotional support actions.
  • Unifies everything into one clear daily plan.

Instead of giving random tips, the system makes sure all suggestions agree with each other and match the user’s current phase and comfort level.

How we built it

We built HerCycle as a multi-agent system with one central coordinator.

  • Specialized Agents: Each agent focuses on one area—cycle patterns, symptoms, nutrition, movement, emotional support, and sustainability.
  • Master Coordinator: Decides which agents to consult and merges their outputs.
  • Safety Layer: Ensures suggestions are non-medical and supportive.
  • Backend ML: A machine learning model runs to predict cycle timing.
  • Privacy: All data is processed locally for the prototype to keep things simple and transparent.

The frontend shows the final unified plan and lets users explore what each agent contributed.

Challenges we ran into

  • Agent Collaboration: Designing agents that actually collaborate instead of repeating the same advice.
  • Managing Uncertainty: Balancing machine learning predictions with uncertainty—periods aren’t perfectly predictable.
  • Safety Constraints: Making sure the system doesn’t sound medical or give unsafe advice.
  • Simplicity vs. Power: Keeping the architecture simple enough to explain, but powerful enough to feel meaningful.

Accomplishments that we're proud of

  • Building a real multi-agent system with clear roles and coordination.
  • Integrating an explainable ML model into an agent workflow.
  • Making recommendations that respect real-world constraints like time, budget, and energy.
  • Creating something that feels thoughtful, not overwhelming.

What we learned

  • Multi-agent systems work best with a strong coordinator and clear responsibilities.
  • In health-related domains, safety and clarity matter more than complexity.
  • Explainability builds trust both for users and for evaluators.
  • Good AI is less about showing intelligence and more about showing care.

What's next for HerCycle – Menstrual Wellness Companion

  • Add long-term trend insights and visual summaries.
  • Improve personalization as more cycle data is logged.
  • Introduce optional premium agents for deeper symptom analysis.
  • Explore voice or regional language support.
  • Partner with local health resources for better access and guidance.

Built With

Share this project:

Updates