## Inspiration
Every month I felt like a completely different person. Different energy, different focus, different skin, different emotional capacity, and I had no idea why. I had five separate apps trying to track pieces of my health and none of them talked to each other or explained anything. The deeper I went into menstrual cycle science, the more I realised my cycle wasn't just a fertility metric. It's a full biological operating system that governs cognition, mood, metabolism, skin, sleep, and physical performance. No app was treating it that way. So I built one that does.
## What it does
Aura is a personal AI wellness OS built entirely around the menstrual cycle. Every morning, a two-minute check-in captures your mood, energy, sleep, and what you need to get done. Aura knows your cycle phase and uses that, combined with peer-reviewed endocrinology, to surface personalised guidance across every dimension of your day. Your macro targets shift by phase. Your hydration goal adjusts because progesterone raises body temperature in the luteal phase. Your workout recommendation changes because muscle protein synthesis peaks with oestrogen. Your skin advice, sleep tips, and PMS toolkit are all grounded in the actual hormonal science for that specific day of your cycle. The AI is completely invisible. No generate buttons, no configuration. It just knows.
## How we built it
React 18, Firebase Firestore with offline persistence, Framer Motion for animations, and a multi-provider AI system routing through Anthropic, OpenAI, Gemini, and Groq with automatic fallback. The cycle intelligence lives in a science library built from scratch from a comprehensive menstrual cycle research document. Every phase has its hormonal profile, cognitive effects, nutritional priorities, fitness guidance, skin behaviour, and supplement evidence written as structured constants the app references directly. User API keys are encrypted with AES-256-GCM using the Web Crypto API before touching the database. The AI proxy runs as a Vercel edge function. The whole thing is a PWA installable from the browser with no app store required.
## Challenges we ran into
Getting the cycle phase intelligence right was harder than expected. Phase thresholds had to be proportional to each user's actual cycle length, not hardcoded day ranges, because a 35-day cycle ovulates on day 19, not day 14. Firebase's even-segment path rule caused silent write failures that took real time to diagnose. Making the AI feel invisible required rethinking every UI touchpoint and removing every label, button, and loading state that made the technology visible rather than the experience. And balancing the emotional tone of a wellness app with hard medical science, without tipping into either generic positivity or clinical coldness, was a constant design challenge throughout.
## Accomplishments that we're proud of
The science library. Every insight in Aura is traceable to actual endocrinology. The connection between progesterone withdrawal and GABA receptor disruption, why cruciferous vegetables support oestrogen clearance, why pain tolerance peaks at ovulation. Writing that as structured data the AI can reference rather than relying on the model to hallucinate wellness advice feels genuinely meaningful. We're also proud of the invisible AI. Users experience personalisation without ever configuring anything.
## What we learned
That the hardest part of building a health app isn't the technology, it's the tone. Every screen has to feel safe enough to be honest in. The PMS toolkit had to acknowledge the actual neuroscience of what's happening without being dismissive or overwhelming. Logging that you're having a rough mental health day requires a completely different emotional register than a fitness tracker asking for your reps. Getting that right required as much thought as any technical decision we made.
## What's next for Aura
Opening to a waitlist. The science library becomes a living document updated with new research. Wearable integration so Oura Ring HRV data feeds directly into the daily phase intelligence. A shared view for partners who want to understand the cycle without it being a conversation the user has to initiate every time. And long-term, pattern recognition across multiple cycles so Aura can tell you not just what the science says about your phase, but what your own data says about how your specific body responds to it.
Log in or sign up for Devpost to join the conversation.