Inspiration
As a student, my days became a blur—late nights, poor hydration, rushed meals, constant stress, and cortisol through the roof. It didn’t just sap my energy; it threw my hormones off, made my periods irregular, and brought a wave of symptoms I didn’t understand. I realized many modern women live this way. Holistic practices (Ayurveda, Chinese medicine, community wisdom) exist, yet they’re scattered. Clinical care often looks at snapshots, not daily patterns. I wanted a grounded, modern tool that blends both worlds.
What it does — Hormone Harmony
Hormone Harmony is an AI-guided wellness app that helps women build steady daily routines for hormonal balance and gives healthcare providers (HCPs) an optional dashboard to see consenting patients’ trends. It supports daily tracking of sleep hours, stress (1–5), hydration cups, meals quality, activity, cycle phase/day, mood (1–5), and symptoms. The AI coach, LovelyLadyLilac, turns metrics into clear daily plans (summary with three tips, two resources, and a gentle disclaimer) and weekly summaries. With consent, assigned HCPs can view trends, read the latest plan, and add secure notes. A cycle calendar and insights view show patterns—for example, low sleep plus high stress preceding irregular cycles. A central chat page keeps all AI interactions in one place for continuity.
Why it matters — Curator’s Cause
Lives: practical, personalized routines that improve day-to-day wellbeing.
Communities: honors culturally rooted practices alongside modern data.
Tech for impact: AI used for access and clarity—not gimmicks.
How I built it
I framed the problem, user stories, and data model, then turned them into concise builder prompts. The no-code app was created in Base44 with multi-role access (member + HCP) and consent-based privacy so HCPs only see assigned patients who opt in. The agent LovelyLadyLilac returns predictable JSON (summary, three tips with “Title | Action | Rationale”, two links, disclaimer) so plans save and render reliably. The layout is responsive and mobile-first, with accessible contrast and a slide-in menu on small screens. Dashboard, Insights, and HCP actions deep-link into /chat, so conversation and plans stay together.
Challenges
Finding the right balance between specific, doable guidance and avoiding medicalized language; enforcing strict, valid JSON from the AI every time; designing consent as the default gate for HCP visibility; removing mobile side-scroll and stacking charts/cards cleanly.
What I’m proud of
A calm, respectful tone that centers women’s real lives; a clean HCP workflow that adds context between visits without replacing care; culturally aware tips that are practical today—warm meals in luteal, breathwork for high stress, and evening wind-down for low sleep.
What I learned
Good prompts behave like product specs when they’re short, exact, and testable. “Consent + assignment” is the simplest, safest way to scope HCP access. For wellness apps, clarity beats complexity; fewer, meaningful metrics drive better adherence.
What’s next
Gentle habit streaks and reminders; simple export of de-identified trends for research; multi-language support and region-aware resources; optional device integrations (sleep/activity) to reduce manual input.
Safety & ethics
Guidance is educational only; no diagnosis or prescriptions. If severe pain, heavy bleeding, fainting, or pregnancy concerns appear, the app suggests seeking clinical attention. Users can change consent at any time.
Built with
Base44 — no-code builder for UI, data, roles, routing, and the chat hub.
Gemini 2.5 Pro — AI model powering the LovelyLadyLilac coach and summaries (configured for strict JSON outputs).
ChatGPT — helped refine product framing, prompts, and copy.
Built With
- and
- and-the-chat-hub.-gemini-2.5-pro-?-ai-model-powering-the-lovelyladylilac-coach-and-summaries-(configured-for-strict-json-outputs).-chatgpt-?-helped-refine-product-framing
- base44-?-no-code-builder-for-ui
- copy.
- data
- prompts
- roles
- routing
Log in or sign up for Devpost to join the conversation.