For a decade, wearables taught people to interpret their own bodies with data.
It worked. It taught a generation to think in HRV, heart rate, strain, and recovery.
Now, my job is to make WHOOP AI native — design how the next generation understands their health.
I identify where AI can make WHOOP more personal and design those opportunities into end-to-end features. My work spans shared interaction patterns and UX primitives, alongside defining the agent logic that guides how and when AI interacts with users. Together, these shape an AI coach that narrates each member’s health story and builds a long-term coaching relationship that adapts to their goals, routines, and everyday life.
The opportunities for personalization emerge from a connected loop: capture context across member touchpoints, interpret it, deliver relevant guidance, and plan for next steps to achieve health goals. Each interaction adds context back to the data layer. As that context grows, individual experiences become part of a richer, more personal health story. That is narrative intelligence.
Context-driven loop
Builds and maintains the context captured in daily interaction — memory that grows with the user.
The same data, a different story — combines biometrics with memory and generates AI charts that belong to you specifically.
A story only matters if it arrives at the right time, with the right message — brought forward when it changes a decision.
Composes a program from what the system knows about the user's behavior, intent, and adherence — adaptive, personalized, and easy to act on.
Within 3 quarters, I led design of AI features across four pillars, and brought
In Q1 2026, I defined the experience framework for WHOOP’s memory system, bringing fragmented personal context into a structured Context Hub centered on members’ health goals. I partnered with AI engineers to define memory categories and weighting rules that guide how Coach prioritizes and uses that context.
I designed around key philosophy of transparency and control, making Coach’s understanding visible and giving them an active role in shaping their personalized coaching experience. Revisiting past interactions also lets members reflect on their health journey holistically.
Field note — 01
But memory alone is just storage. How the system going to use it and make it valuable to members?
Memory helps Coach interpret health data in the context of members’ lives. Charts show what changed, but members often have to work out why it matters. I designed an AI context layer on top of WHOOP’s charts, connecting biometrics with personal context to explain patterns and possible contributing factors.
The layer clearly distinguishes AI interpretations from measured data, with each interpretation’s prominence reflecting confidence in the connection.
01 · Recovery trend
02 · Live stress
Field note — 02
Insight that never reaches you at the right moment isn't insight, it's a file in database.
As Coach gains more context, the challenge is deciding what deserves a member’s attention now. Daily Loop is a context-aware pop-up every morning. I designed Daily Loop’s agent logic, widget library and worked with the team to define prioritization logic, helping the agent select one meaningful takeaway based on the time of day, current data, goals, and personal circumstances. Each insight builds on previous guidance and member input, keeping daily coaching concise, relevant, and continuous.
Field note — 03
And how might we turn daily insights into actions, to add new years to your life?
Program is the most important strategic project in 2026. I designed end to end experience to turn members’ intentions into actionable daily plans that adapt as their capability and circumstances change. I distilled an open-ended range of goals into core types, shaping page narratives and modular components that make each plan consistent, easy to follow, and motivating, with clear progress and next steps.
With AI engineers, I set boundaries for plan generation and designed when Coach should adjust plans or check in, applying behavior-change principles to help members realistically complete.
Your journey becomes memory
Day-to-day coaching can’t be designed as a fixed set of screens. I built a flexible, platformized system of modular components that the agent can assemble in real time, based on each user’s context, needs, and moment. The interface and chatbot is only the surface; behind it is the agent logic I designed, deciding when to step in, what matters, and how to deliver it.
Weekly targets
Intent capture
Event
Summary strip
Planned activity
Tips list
Commit card
Strength card
Metric stack
Trend chart
Activity strain
Check-in flow
Smart alarm
Reminder
Section header
Activity row
Insight banner
Highlight
Routine builder
Stacked chart
Header
Range slider
Impact barAgent flow behind the screen
Learn from experiments, and learn from users' natural behavior. Shipped surfaces are the real research instrument.
Balance flexibility, scalability, and optimal UX. Every generated surface has to hold up across thousands of contexts you'll never see.
Long-term companionship and accumulated data beat any one-off experience — the design of the touchpoints matters most.
