femAI

Inspiration

Menstrual wellness is often tracked across notes, calendars, and disconnected apps. We wanted to create one calm, privacy-conscious space where users can understand patterns in their cycle, mood, symptoms, and energy while receiving supportive AI guidance.

What it does

FemAI lets users register securely, track cycles, log symptoms and moods, complete daily check-ins, view summaries, chat with an AI wellness coach, manage privacy settings, and export or delete their data. It also includes partner-sharing foundations, notifications, admin insights, and a Sepolia testnet Premium payment flow.

How we built it

We built FemAI with React and Vite on the frontend, NestJS and TypeScript on the backend, PostgreSQL with Prisma for persistence, Redis for rate limiting and caching, JWT authentication, and OpenAI-compatible AI integration. Payments use Ethereum Sepolia verification with browser-wallet support.

Challenges we ran into

The biggest challenges were connecting multiple services reliably, managing Prisma schema synchronization, handling Windows development environment issues, and safely designing AI responses for a health-related product. We also had to ensure payment verification checks the correct recipient, amount, chain, and transaction status.

Accomplishments that we're proud of

We created a working full-stack demo with authentication, live database persistence, AI chat, cycle and wellness tracking, privacy controls, and blockchain payment foundations. The app runs locally with PostgreSQL, Redis, frontend, and backend connected.

What we learned

We learned that health-focused AI needs strong safety boundaries, clear disclaimers, and privacy-first design. We also learned how important end-to-end verification is when combining frontend, backend, databases, AI providers, and blockchain services.

What's next for FemAI

Next, we plan to add richer visual insights, real notification delivery, completed partner invitation flows, production-ready Prisma migrations, expanded testing, secure deployment, and broader accessibility support.

Built With

• chatgpt
• claude
• codex
• fable5
• openrouter
• typescript
• vscode