What this explores
A set of explorations focused on how AI can transform health record management from static data storage to intelligent, assistive experiences that help users make better decisions about their health.
Key Directions
1. Smart Summaries
AI-generated summaries that simplify complex reports into clear, easy-to-understand insights—highlighting what actually matters.
2. Contextual Insights
Surface patterns across reports (e.g., trends in vitals, recurring conditions) to help users understand their health over time.
3. Conversational Access
Chat-based interface to query health records—“What changed in my last report?” or “Show my recent prescriptions.”
4. Proactive Alerts
Timely nudges for abnormal values, missed medications, or upcoming tests—turning passive data into active care.
5. Guided Actions
Suggest next steps like booking tests, consulting doctors, or sharing reports—reducing decision friction.
Design Principles
Clarity over completeness → Show what matters, not everything
Trust through transparency → Explain AI decisions clearly
Privacy by design → Give users full control over their data
Assist, don’t overwhelm → AI as support, not noise
Outcome Vision
From static records → to an intelligent health companion
From information overload → to clear, actionable guidance
The health records no one understood
Every report was there.
Blood tests, prescriptions, scans—neatly uploaded, properly labeled, technically accessible.
And yet, most users didn’t really use their health records.
They opened them when needed.
Scanned for familiar words.
Maybe shared them with a doctor.
But understanding? Acting? Connecting the dots?
That rarely happened.
The system stored information.
It didn’t create clarity.
The comprehension gap
Health records aren’t just data.
They carry decisions, anxiety, and responsibility.
But most systems treat them like files:
PDFs stacked in folders
Medical jargon without context
Numbers without meaning
For users, this creates friction:
Is this normal?
Has this changed over time?
Do I need to act on this?
Designers saw organization.
Users felt confusion.
From storage to intelligence
We stopped asking:
“How do we organize records better?”
And started asking:
“How do we help people understand their health?”
The answer wasn’t better folders.
It was better interpretation.
AI became the layer that bridges data and understanding:
Translating complex reports into simple summaries
Highlighting what’s important—and what’s not
Connecting patterns across time, not just single reports
The system started thinking with the user.
Designing for clarity and trust
Health is sensitive.
So AI couldn’t feel like a black box.
Every insight needed to be:
Explainable → Why is this flagged?
Contextual → How does this compare to before?
Actionable → What should I do next?
We introduced:
Smart summaries for quick understanding
Trend views to show changes over time
Guided actions like booking tests or consulting doctors
The goal wasn’t to replace doctors.
It was to reduce confusion before reaching one.
From passive records to active guidance
Instead of waiting for users to dig through data,
the system started surfacing what mattered:
Abnormal values highlighted proactively
Medication reminders triggered at the right time
Relevant reports surfaced during consultations
The experience shifted from reactive to assistive.
Users didn’t just store their health data.
They started engaging with it.
What changed
The transformation wasn’t visual—it was behavioral.
Expected impact:
+35% increase in report interaction time
+28% improvement in user understanding (self-reported clarity)
+22% increase in follow-up actions (consults, tests, sharing)
-30% reduction in repeated confusion-driven support queries
Users moved from asking:
“What does this mean?”
To:
“What should I do next?”
From records to relationships
Before, health records were static.
After, they became a living system—one that helps users understand, track, and act on their health over time.
Because in healthcare,
clarity isn’t just a better experience.
It’s a better outcome.
