top of page
2024-2025
AiTuki – Turning Cortisol Data Into Just-In-Time Actions
From Perimenopause Niche to Cortisol-Smart Stress Partner
About AiTuki
Prototype Demo
Duration
6 months
Team
12 members
My Role
Product Designer
Tools
Figma, Miro, GPT
Project Overview
When I joined AiTuki, the idea was heartfelt but narrow: help women in perimenopause navigate hormone chaos with an app. It was clear there was pain—hot flashes, broken sleep, and brain fog—but as we listened to more people, a bigger pattern emerged.
Men, younger women, and high-performing professionals were describing the same exhaustion, the same "wired-but-tired" feeling. The common thread wasn't a life stage. It was chronic metabolic stress driven by cortisol.
Problem and Market Gap
Existing wellness apps flood knowledge workers (30–55, 60% women) with raw wearable data—cortisol spikes, sleep scores, and energy crashes—but provide no actionable next steps.
Users said, "I want something that tells me when to stop and what small thing to do now."
Millions
Hormone/metabolic issues
80M
Chronic stress cases
40M
Insomnia sufferers
To move from intuition to evidence, I led a focus group and a survey study with 40 knowledge workers aged roughly 30–55, most of whom were already using wearables or health apps. We explored how they currently manage stress, their feelings about hormones, and what they would actually pay for.
Language testing was even more revealing. The "Perimenopause app" made some women feel seen, but it also made others feel boxed in, and it completely shut out men and younger users. "Metabolic stress" and "cortisol" landed better; people recognized themselves in symptoms like burnout, insomnia, and stubborn weight, even if they hadn't used the terms before.
The Pivot
Perimenopause moved from being the market to being one powerful use case inside a broader metabolic-stress platform.
The numbers and quotes told a clear story:
Were comfortable paying $5–10 per month if the app helped in the moment, not just tracked.
Said they'd prepay 6 months to avoid "yet another tracking app."
(15 people) immediately volunteered to join our Cohort-1 cortisol case study, with 24 more on the waitlist.
User Research and the Pivot
Cortisol became the backbone of the MVP for both human and business reasons. Biologically, it sits upstream of many complaints users shared, including racing thoughts at night, energy crashes, stubborn fat, and mood swings. Strategically, it gave us a simple story:
"If we help you understand and nudge your cortisol patterns, we can improve sleep, energy, and focus—not just one symptom."
Choosing cortisol made the product easier to explain to users and investors, while still leaving room to expand into other hormones later.
MVP Focus:
Linking wearable data and self-reports to cortisol-related patterns.
Delivering 1–2 Just-In-Time Adaptive Interventions (JITAI) per day—micro-walks, bedtime shifts, hydration prompts, and "time to pause," each with a short "why this now" explanation.
Why is Cortisol the First Hormone
Building a cortisol-smart AI isn't just a design problem — it's a data problem. To give users meaningful, just-in-time nudges, our AI needed to learn from real habit patterns linked to real clinical outcomes. That data simply didn't exist off the shelf.
Step 1 — Collect
Using Movement.so, we built structured habit-tracking programmes and enrolled early cohort users. The app captured daily physical, emotional, mental, and energy data — steps, sleep, stress scales, fatigue, screen time, and more — over 8-week windows.
Step 2 — Annotate
Partnering with specialist doctors, we layered clinical recommendations on top of each tracked pattern. Physicians reviewed anonymised user data and mapped cortisol-relevant signals to specific interventions — creating ground-truth labels for the model.
Step 3 — Train
With labelled habit data and doctor-validated recommendations in hand, we trained AiTuki's AI to predict the right nudge at the right moment — not based on averages, but on each user's own longitudinal stress signature.
This pipeline — user tracking → doctor annotation → model training — is what gives AiTuki its clinical credibility. The JITAI nudges aren't guesses; they're grounded in evidence collected from real people and validated by real clinicians.
Data Collected via Movement.so
These screens show the breadth of data tracked across the four pillars — Physical, Emotional, Mental, and Energy — that fed directly into the AI training pipeline.
The Hidden Challenge: Training the AI
Our breakthrough concept became the "Smart Mentor," a friendly, adaptive guide that learns your stress rhythms and suggests 1–2 personalised actions daily. No dashboards. No guilt notifications. Just meaningful, just-in-time micro-interventions.
And I worked alongside Manasa, our UI designer, who translated those flows into beautifully calm, inclusive visuals. Together, we created an interface that feels like a breath of air. Simple gradients, soft motion cues, and typography that communicates rest rather than rush.
We built a working prototype to validate the cortisol-first concept with real users.
Try the live product built during the research phase →
Try the Demo
Branding, Colour & Typography
Colour System
Typography Scale
Component Library
Design System Theme
AiTuki's calm, mentor-like design system uses soft teal gradients, accessible Inter typography, and subtle state-based colours to evoke stress relief rather than clinical tracking.
Primary
#00BCD4 Teal
Typography
Inter (Variable)
H1 34px → Caption Cell
Style
Soft gradients · Calm motion · Wave forms
As the lead product designer, I defined:
The core product strategy and MVP roadmap.
Research synthesis and persona storytelling.
End-to-end UX flows, from onboarding to adaptive daily actions.
Motion design system and narrative product video for fundraising.
Solution Design: From Dashboards to Decisions
User Impact
✓ Users reported feeling "guided, not judged" by the 1–2 daily actions and appreciated the short explanations tied to their own patterns.
✓ Perimenopause remained meaningfully supported but no longer defined or limited the product—users could see themselves in the story regardless of gender or life stage.
Business Impact
✓ Expanded TAM from a perimenopause-only niche to a broad metabolic stress market (tens of millions with chronic stress and insomnia), improving long-term revenue and partnership potential.
✓ Generated strong early demand signals: 37.5% cohort conversion, high willingness to pay, and a clear, defensible positioning vs. single-symptom apps.
✓ Gave founders a crisp investor narrative: AiTuki as the decision layer on top of wearables, not another data tracker.
40+
User Interviews
83%
Willing to Pay
37.5%
Cohort Conversion
Impact and Outcomes
Key Takeaways
Human-Centered Design
The calm, mentor-like interface and language made users feel "guided, not judged," fostering trust and engagement.
Just-In-Time Actions
Moving from passive tracking to active guidance created a differentiated product that users were willing to pay for.
Strategic Focus
Starting with cortisol as the first hormone created a simple, compelling narrative while leaving room for future expansion.
Research-Driven Pivot
User research revealed that the real pain point wasn't perimenopause specifically, but metabolic stress affecting a much broader audience.
bottom of page