health-tech start up
Auri App
role
PRODUCT
End-to-End
Product
tools
Figma, FigJam, Claude, ChatGPT, Lyssna, Google Suite, Stitch AI, Otter AI
when
June 2025 (3 weeks)
People are trapped in a cycle of catching stress after its too late.
Problem
People only recognize stress after it overwhelms them β users remain trapped in a reactive cycle of damage control instead of prevention.
Solution
Auri uses AI-powered biometric pattern recognition to catch stress before it spirals and deliver immediate intervention.
Constraints
3-week sprint: Used agile methods and AI tools to prototype and iterate rapidly.
Hyper-selective user testing: Engaged ~10 users (combining interviews and testing) who closely matched the target audienceβthose facing anxiety or high stress, both wearable users and not.
Learning in motion: Adapted quickly to new AI-based design tools for rapid experimentation and validation.
Users lack awareness of stress onsetβthey only recognize it after overwhelming
There is a need for predictive intervention before spiral begins, not reactive management
Users don't want to do tedious work around stress management or their health, they're stressed enough!
When users do have their biometric data, they don't know what to do with it.
Learning from Oura Ring & Natural Cycles
β How might we transform stress management from reactive damage control to effortless, proactive prevention?
Design Principles
Effortless
Predictions happen automatically & Stress moments require minimal cognitive load
Trustworthy
Scientifically valid stress detection, not arbitrary metrics
Proactive
Prevent stress episodes rather than manage aftermath
Key Differentiators:
Zero manual logging required
β They need sustainable solutions that fit chaotic lifestyles.
Biometric-driven personalization
β They need in-the-moment help, not generic advice
Real-time intervention delivery
β Information without action leads to abandonment
Core Concept: Utilize AI to analyze biometric patterns and deliver personalized stress support before things escalate.
Why? Based on our interviews & formative research, AI is a tool to be used to analyze biometrics and predict stress, helping to raise awareness in users and for the app to learn about the user.
Turning user stories into personas to keep empathy at the core of design
Creating a feature matrix to analyze the imapct it will have for our users.
Maya & Jordan's Journey through building stress resilience
Blueprinting User Navigation
Creating the building blocks of the app that are will fulfill Maya and Jordan's tasks to manage and discover their stress for the MVP.
AI-Accelerated Design Workflow
1st Iteration: Smart AI integration under a 3-week sprint
To challenge myself under a 3-week constraint, I leveraged AI not just for speedβbut to deepen the quality of design thinking without compromising clinical rigor.
Stitch AI for UI:
Cut wireframing time by 50%, enabling rapid user feedback cycles from sketches to testable prototypes.
*Helped me decide to pause the wellness library based on feedback.
The Impact: This approach compressed weeks of research synthesis into days while ensuring clinical accuracy. Instead of arbitrary stress metrics, Auri maps physiological patternsβelevated HR, disrupted sleep, shallow breathingβto established psychological research.
Low to mid-fidelity Design:
sketches to concept wireframes
Strategic Design Risk: Dark Mode
I chose dark mode to position Auri as "luxury wellness within reach"βsimilar to Oura Ring's premium approach. In a market of bright, clinical interfaces, dark mode reduces overstimulation while signaling sophistication, making proactive stress management feel both elevated and accessible.
The key words driving the design of the product are:
modern / organic / intelligent / elevated
mid-fidelity to high-fi's
Users loved the proactive approach, biometric validation, and effortless experience - but there were a couple refinements to be done.
User Testing Approach Conducted qualitative study with 6 participants matching target criteria (anxiety patterns, mixed wearable usage).
Three Core Tasks:
Navigate from stress prediction notification to coping tool to biometric feedback
Switch coping tools during busy moment from homepage
Extract meaningful insights from biometric data and take action
The overall feedback was very positive. The iteration feedback focused on clarity (stress scores, navigation) and making the experience feel less assumptive about users' emotional states.
Iteration Priorities:
Stress Score Clarity:
Enhanced visualization and explanation to help users immediately understand what their biometric stress levels mean and why they matter
UX Language Refinement:
Simplified terminology and messaging to reduce cognitive load during stress moments, making interactions feel supportive rather than clinical
Usability Enhancement:
Added user control through quick-select options, switch features, and feedback mechanisms (thumbs up/down) to create more efficient, personalized interactions
Iterations, ALL FROM
INCORPORATING USER FEEDBACK
Success Metrics: Qualitative Validation
Validated User Attitudes: Users were excited about catching stress before they spiral. The solution addresses the core frustrationβlacking awareness of stress onset until it's too late to prevent overwhelm.
π¬ Adam, User Testing Participant
Real Quote from the Usability testing
Final flows
What I Learned
If I Did This Again:
Run competitive analysis earlier
I should have mapped AI chat tools (Claude, ChatGPT, Headspace) upfront to define chatbot differentiation before prototyping. Would have sharpened intervention strategies and avoided mid-sprint pivots.
Apply proven patterns more intentionally
I reinvented wheels during prototyping when established interaction patterns existed. Leveraging design systems earlier would have preserved creative energy for novel problems.