Inspiration HerSync was born from the loss of a close female relative to a delayed health diagnosis. This tragedy, combined with our passion for cutting-edge tech, inspired us to build a solution that delivers early, accessible, and culturally-aware care for women, especially in Africa where gaps in reproductive and diagnostic access persist.
What it does HerSync is an AI-powered cycle tracker and multilingual reproductive health assistant that works both online and offline. It helps users:
Log periods and symptoms Predict ovulation and cycle patterns Detect potential irregularities Access reproductive health tips in local languages Connect to care through WhatsApp or the mobile app
All while protecting privacy and working in low-connectivity settings.
How we built it We combined: Flutter for mobile app development Firebase (online sync, fallback storage) Gemini AI for health conversation and support TinyML to run lightweight ML models offline on-device LSTM for cycle prediction Isolation Forest for anomaly detection Twilio for WhatsApp integration NSCache / SQLite for local persistence
Our models are optimized to run in low-resource environments, making HerSync usable without constant internet access.
Challenges we ran into Training accurate ML models that work on-device (TinyML) Ensuring health advice is safe, relevant, and localized Maintaining user privacy without cloud dependency Adapting the UI for users with low digital literacy Navigating fragmented health data policies across regions Accomplishments that we're proud of Successfully deployed offline cycle prediction using TinyML Early adoption by users in Nigeria, Ghana, and rural UK areas 90%+ engagement retention in low-bandwidth settings Support for Yoruba and Pidgin, with Arabic and French in progress Shortlisted for multiple health-tech grants and accelerators
What we learned Offline-first AI is not just helpful – it’s necessary in Africa Users trust products that speak their language and respect their culture Privacy and explainability are as important as accuracy in health AI Building with empathy improves adoption and impact
What's next for HerSync Expand offline diagnostics (e.g., early signs of PCOS, anemia) Integrate wearable syncing with low-cost health devices, leveraging IoT Launch pilot with schools and health centers in West Africa Build partnerships with local NGOs and health ministries Publish open-source datasets and TinyML models for femtech research
Log in or sign up for Devpost to join the conversation.