Epic Case Study
Scaling an AI Design System Across 80+ Clinical Apps
Scaling an AI Design System Across 80+ Clinical Apps
Company
Role
Lead UX Designer
Work
UX strategy & design, design system, visioning
Year
2022 - 2025
Disclaimer: Due to IP restrictions, this case study only includes conceptual visuals. Actual Epic designs and assets cannot be shared, unless they've been made public.
Disclaimer: Due to IP restrictions, this case study only includes conceptual visuals.
Actual Epic designs and assets cannot be shared, unless they've been made public.
Background
Epic is the largest EHR provider in the U.S., serving over 325 million patients across 2,600+ hospitals and has 80+ clinical applications. As generative AI entered the healthcare space, Epic launched dozens of new features - summarization, automation, message drafting - at unprecedented speed. Learn more on Epic’s website.
To keep up, they needed a system: consistent UX, reusable design patterns, and a clear visual language to make AI trustworthy and intuitive across high-stakes tools.
Background
Epic is the largest EHR provider in the U.S., serving over 325 million patients across 2,600+ hospitals and has 80+ clinical applications. As generative AI entered the healthcare space, Epic launched dozens of new features - summarization, automation, message drafting - at unprecedented speed. Learn more on Epic’s website.
To keep up, they needed a system: consistent UX, reusable design patterns, and a clear visual language to make AI trustworthy and intuitive across high-stakes tools.
Background
Epic is the largest EHR provider in the U.S., serving over 325 million patients across 2,600+ hospitals and has 80+ clinical applications.
As generative AI entered the healthcare space, Epic launched dozens of new features - summarization, automation, message drafting - at unprecedented speed. Learn more on Epic’s website.
To keep up, they needed a system: consistent UX, reusable design patterns, and a clear visual language to make AI trustworthy and intuitive across high-stakes tools.
Background
Epic is the largest EHR provider in the U.S., serving over 325 million patients across 2,600+ hospitals and has 80+ clinical applications. As generative AI entered the healthcare space, Epic launched dozens of new features - summarization, automation, message drafting - at unprecedented speed. Learn more on Epic’s website.
To keep up, they needed a system: consistent UX, reusable design patterns, and a clear visual language to make AI trustworthy and intuitive across high-stakes tools.
The Challenge
Each of Epic’s 80+ apps was building AI independently, with different scopes, UIs, and timelines. The result: fragmented experiences, unclear interaction rules, and no scalable way to guide users through AI-driven workflows.
Clinicians didn’t know what to trust. Designers had no shared foundation. Developers lacked reusable components. Adoption was stalling. Without alignment, speed was becoming a liability.
The Challenge
Each of Epic’s 80+ apps was building AI independently, with different scopes, UIs, and timelines. The result: fragmented experiences, unclear interaction rules, and no scalable way to guide users through AI-driven workflows.
Clinicians didn’t know what to trust. Designers had no shared foundation. Developers lacked reusable components. Adoption was stalling. Without alignment, speed was becoming a liability.
The Challenge
Each of Epic’s 80+ apps was building AI independently, with different scopes, UIs, and timelines.
The result: fragmented experiences, unclear interaction rules, and no scalable way to guide users through AI-driven workflows.
Clinicians didn’t know what to trust. Designers had no shared foundation. Developers lacked reusable components. Adoption was stalling.
Without alignment, speed was becoming a liability.
The Challenge
Each of Epic’s 80+ apps was building AI independently, with different scopes, UIs, and timelines. The result: fragmented experiences, unclear interaction rules, and no scalable way to guide users through AI-driven workflows.
Clinicians didn’t know what to trust. Designers had no shared foundation. Developers lacked reusable components. Adoption was stalling. Without alignment, speed was becoming a liability.
My Role
I co-led the design and scale-up of Epic’s AI UX system - from pattern audits to system creation to cross-team rollout.
I helped define the core interaction types (e.g. summarization, task automation), built out reusable UX and UI patterns, created a visual system to flag AI-generated content, and equipped teams with guidance, documentation, and guardrails.
This became the UX foundation for all AI-powered experiences at Epic, trusted by 80K+ clinicians and deployed across 330+ organizations.
My Role
I co-led the design and scale-up of Epic’s AI UX system - from pattern audits to system creation to cross-team rollout.
I helped define the core interaction types (e.g. summarization, task automation), built out reusable UX and UI patterns, created a visual system to flag AI-generated content, and equipped teams with guidance, documentation, and guardrails.
This became the UX foundation for all AI-powered experiences at Epic, trusted by 80K+ clinicians and deployed across 330+ organizations.
My Role
I co-led the design and scale-up of Epic’s AI UX system - from pattern audits to system creation to cross-team rollout.
I helped define the core interaction types (e.g. summarization, task automation), built out reusable UX and UI patterns, created a visual system to flag AI-generated content, and equipped teams with guidance, documentation, and guardrails.
This became the UX foundation for all AI-powered experiences at Epic, trusted by 80K+ clinicians and deployed across 330+ organizations.
My Role
I co-led the design and scale-up of Epic’s AI UX system - from pattern audits to system creation to cross-team rollout.
I helped define the core interaction types (e.g. summarization, task automation), built out reusable UX and UI patterns, created a visual system to flag AI-generated content, and equipped teams with guidance, documentation, and guardrails.
This became the UX foundation for all AI-powered experiences at Epic, trusted by 80K+ clinicians and deployed across 330+ organizations.
Disclaimer: Due to IP restrictions, this case study only includes conceptual visuals. Actual Epic designs and assets cannot be shared, unless they've been made public.
📢 Full Case Study In Construction!
📢 Full Case Study In Construction!
A deeper dive into this case study is in the works, but if you want to hear about it now, let’s chat!
A deeper dive into this case study is in the works, but if you want to hear about it now, let’s chat!
When Speed Outpaces Structure
Epic jumped into generative AI within months of ChatGPT’s release. In partnership with Microsoft, we launched 100+ AI-powered features across chart summaries, documentation, patient messaging, and automation.
When Speed Outpaces Structure
Epic jumped into generative AI within months of ChatGPT’s release.
In partnership with Microsoft, we launched 100+ AI-powered features across chart summaries, documentation, patient messaging, and automation.
But with 80+ apps building AI in parallel, things quickly got messy. UX patterns were inconsistent. Visual cues didn’t align. There were no shared patterns guiding how AI should show up.
Developers reused outdated components. Designers couldn’t be everywhere. Some features looked polished, others felt disconnected.
That fragmentation introduced real risk:
Clinicians couldn’t always tell what was AI-generated, hurting trust
Inconsistencies slowed adoption
Shared design couldn’t scale without cohesion
And poor UX could directly affect patient care
Leadership prioritized speed. But without structure, that speed could backfire. We needed a system, and we needed it fast.
But with 80+ clinical applications, each building AI independently, the result was fragmented experiences. Features launched with inconsistent UI, conflicting visual cues, and no shared UX patterns.
Developers reused outdated components. Designers couldn’t support every project. Some apps had strong AI visibility; others felt disconnected.
But with 80+ clinical applications, each building AI independently, the result was fragmented experiences.
Features launched with inconsistent UI, conflicting visual cues, and no shared UX patterns.
Developers reused outdated components. Designers couldn’t support every project. Some apps had strong AI visibility; others felt disconnected.
This posed serious risks:
Trust could erode if clinicians couldn’t tell what was AI-generated
Inconsistent experiences could stall adoption
Reusability and speed would collapse without design cohesion
Most importantly, poor UX could directly impact patient safety
Leadership prioritized speed. But without structure, that speed would backfire. We needed a system - fast.
This posed serious risks:
Trust could erode if clinicians couldn’t tell what was AI-generated
Inconsistent experiences could stall adoption
Reusability and speed would collapse without design cohesion
Most importantly, poor UX could directly impact patient safety
Leadership prioritized speed. But without structure, that speed would backfire. We needed a system - fast.
Our alarms went off - "this could really hurt scalability and adoption". So, me and another UX Designer took initiative to start bringing order to what was quickly becoming a fragmented AI experience across 80+ apps.
First, we grounded ourselves in research.
Creating Epic’s AI Design System
This wasn’t a top-down initiative. It started from the ground.
While working on separate AI projects, a few of us designers noticed the same thing: visuals were clashing, interactions weren’t aligned, and trust signals were all over the place.
We knew that kind of inconsistency wouldn’t scale. So another UX designer and I took the lead and began building structure around what was quickly becoming a fragmented AI experience.
Creating Epic’s AI Design System
This wasn’t a top-down directive - it started from the ground.
While working on separate AI projects, a handful of us designers spotted major inconsistencies in how AI features were starting to look and feel.
Creating Epic’s AI Design System
This wasn’t a top-down directive - it started from the ground.
While working on separate AI projects, a handful of us designers spotted major inconsistencies in how AI features were starting to look and feel.
Our alarms went off - "this could really hurt scalability and adoption". So, me and another UX Designer took initiative to start bringing order to what was quickly becoming a fragmented AI experience across 80+ apps.
First, we grounded ourselves in research.
How We Grounded the Work
Speed was critical, so our research had to be sharp and scrappy. We focused on validating assumptions and finding patterns early:
Audited live and in-progress AI features
Ran pattern-mapping workshops to surface overlaps and gaps
Held interviews with customers and internal SMEs
Looked at how top AI tools handled similar problems
Pulled insights from past predictive UX work and ML research
The goal was to build a strong foundation that scaled proactively, not reactively.
The 4 Core AI UX Patterns
To simplify the chaos, we categorized features into four core UX patterns:
Summarization
Chart digests, medication updates, discharge insights.
Drafted Text
Messages, documentation starters, denial letters.
Transformed Content
Language simplification, conversational queries, data reshaping.
Task Automation
Follow-ups, billing suggestions, prior auths.
Summarization
Chart digests, medication updates, discharge insights.
Drafted Text
Messages, documentation starters, denial letters.
Transformed Content
Language simplification, conversational queries, data reshaping.
Task Automation
Follow-ups, billing suggestions, prior auths.
For each, we designed reusable components:
Conversational UI (textboxes, threads, message swaps)
Summary cards for structured insights
Field styling to distinguish AI vs. human inputs
Floating text editors for easy edits in context
Buttons, banners, loaders, and error states for clarity
Hover states and layered interactions for explainability
Trust-focused elements like feedback prompts and validation steps
Note: Due to IP restrictions, I can’t show this work, but some designs are visible in public.
Summarization
Chart digests, medication updates, discharge insights.
Drafted Text
Messages, documentation starters, denial letters.
Transformed Content
Language simplification, conversational queries, data reshaping.
Task Automation
Follow-ups, billing suggestions, prior auths.
Visual Language and Brand
Epic’s core design system - Hyperspace - wasn’t built for generative AI. We needed a new visual identity that felt distinct, but still aligned with the rest of the ecosystem.
At the center of that identity was the “Bloom” icon, which I helped name and define as the signal for gen AI across all of Epic.
We also created:
Visual Language and Brand
Epic’s core design system - Hyperspace - wasn’t built for generative AI. We needed a new visual identity that felt distinct, but still aligned with the rest of the ecosystem.
At the center of that identity was the “Bloom” icon, which I helped name and define as the signal for gen AI across all of Epic.
We also created:
A new AI-specific icon system
A distinct color palette and gradient system
A distinct color and gradient system
Visual motifs like swoops and spirals
Naming and branding conventions
A new AI-specific icon system
A distinct color and gradient system
Visual motifs like swoops and spirals
Naming and branding conventions
Our goal was to make AI feel powerful, safe, and well-integrated - not bolted on.
Rolling It Out
Once the system felt solid, we validated it with continuous feedback and usability testing. Clinicians responded positively.
And when we brought it to the C-suite, the support was immediate - they saw the strategic value, and in their words, “how beautiful it is.” That greenlight let us move fast:
Rolling It Out
Once the system felt solid, we validated it with continuous feedback and usability testing. Clinicians responded positively.
And when we brought it to the C-suite, the support was immediate - they saw the strategic value, and in their words, “how beautiful it is.”
That greenlight let us move fast:
Partnered with dev leadership to turn patterns into reusable code
Updated internal UX standards with clear, practical guidance
Built a Figma component library so designers could move faster with consistency
Delivered educational presentations to over 4,000 devs, QMs, and designers
Partnered with dev leadership to turn patterns into reusable code
Updated internal UX standards with clear, practical guidance
Built a Figma component library so designers could move faster with consistency
Delivered educational presentations to over 4,000 devs, QMs, and designers
We timed the rollout just before Epic’s largest annual conference, where 40,000+ healthcare professionals were watching. That launch made a strong first impression, and it worked. Customers praised the visual polish and clarity.
From there, we scaled, and I became the system’s sole owner.
We timed the rollout just before Epic’s largest annual conference, where 40,000+ healthcare professionals were watching.
That launch made a strong first impression, and it worked. Customers praised the visual polish and clarity.
From there, we scaled, and I became the system’s sole owner.
Driving Cohesion at Scale
Taking over the system meant owning strategy, direction, and execution.
I became the point person for anything related to AI UX. My day-to-day covered a lot:
Creating and maintaining scalable UX patterns across 80+ apps
Partnering with dev to build reusable components
Reviewing nearly every AI feature to ensure clarity and consistency
Educating and onboarding designers, QMs, and developers
But the role quickly expanded into a bigger strategic role - partnering with C-suite leaders to shape strategy and how Epic positioned AI UX as a differentiator.
Defined AI narratives for conferences and internal messaging
Helped craft pitches and partnership decks, including with Microsoft
Worked directly with orgs like Mayo Clinic to share insights and align approaches
Built north star maps and vision frameworks for AI workflows
I was at the center of both execution and long-term UX strategy.
Owning the System: Process and Practice
As AI dev accelerated, Epic moved so fast it broke its quarterly release model.
I had to design a UX process that kept up, without compromising quality.
Most patterns followed this rhythm:
This process made sure every pattern was intuitive, usable, and trustworthy - for teams and end users alike.
1
Research
Audits, SME input, past research, user interviews when needed
1
Research
Audits, SME input, past research, user interviews when needed
2
Define
What are we solving? What pattern does it align with? What’s the risk?
2
Define
What are we solving? What pattern does it align with? What’s the risk?
3
Co-Design
Pulled in designers from other apps to make sure it scaled
3
Co-Design
Pulled in designers from other apps to make sure it scaled
4
Validation
Built and tested prototypes, refined with clinician feedback
4
Validation
Built and tested prototypes, refined with clinician feedback
5
Enablement
Shipped reusable components, updated guidance, trained teams
5
Enablement
Shipped reusable components, updated guidance, trained teams
Driving Cohesion at Scale
Taking over the system meant owning strategy, direction, and execution. I became the point person for anything related to AI UX. My day-to-day covered a lot:
Creating and maintaining scalable UX patterns across 80+ apps
Partnering with dev to build reusable components
Reviewing nearly every AI feature to ensure clarity and consistency
Educating and onboarding designers, QMs, and developers
But the role quickly expanded into a bigger strategic role - partnering with C-suite leaders to shape strategy and how Epic positioned AI UX as a differentiator.
Defined AI narratives for conferences and internal messaging
Helped craft pitches and partnership decks, including with Microsoft
Worked directly with orgs like Mayo Clinic to share insights and align approaches
Built north star maps and vision frameworks for AI workflows
I was at the center of both execution and long-term UX strategy.
Owning the System: Process and Practice
As AI dev accelerated, Epic moved so fast it broke its quarterly release model. I had to design a UX process that kept up, without compromising quality.
Most patterns followed this rhythm:
1
Research
Audits, SME input, past research, user interviews when needed
1
Research
Audits, SME input, past research, user interviews when needed
1
Research
Audits, SME input, past research, user interviews when needed
2
Define
What are we solving? What pattern does it align with? What’s the risk?
2
Define
What are we solving? What pattern does it align with? What’s the risk?
2
Define
What are we solving? What pattern does it align with? What’s the risk?
3
Co-Design
Pulled in designers from other apps to make sure it scaled
3
Co-Design
Pulled in designers from other apps to make sure it scaled
3
Co-Design
Pulled in designers from other apps to make sure it scaled
4
Validation
Built and tested prototypes, refined with clinician feedback
4
Validation
Built and tested prototypes, refined with clinician feedback
4
Validation
Built and tested prototypes, refined with clinician feedback
5
Enablement
Shipped reusable components, updated guidance, trained teams
5
Enablement
Shipped reusable components, updated guidance, trained teams
5
Enablement
Shipped reusable components, updated guidance, trained teams
This process made sure every pattern was intuitive, usable, and trustworthy - for teams and end users alike.
Solving Enterprise-Scale UX Challenges
These were three of the most complex and impactful problem spaces I led - where the right design call shaped adoption, trust, and safety.
Reinforcing Safety: Automation Bias
Automation bias happens when users over-trust AI, even when it’s wrong. In healthcare, that can lead to real consequences.
I partnered with Epic’s ethical ML group and led a multi-month initiative to tackle this head-on. I created a framework teams could actually use: the Risk × Complexity Matrix.
It helped teams answer: How cautious do we need to be here?
The matrix helped teams answer: “How cautious do we need to be here?” It laid out four UX zones:
We broke features into four zones:
Low Risk, Low Complexity – Keep it fast. Let users know AI was used, but don’t slow them down.
Low Risk, High Complexity – Guide the user clearly without adding noise.
High Risk, Low Complexity – Surface the evidence. Show where the AI output came from.
High Risk, High Complexity – Add focus and friction. Support safe, confident decisions.
Low Risk, Low Complexity
Keep it efficient and low-friction - users just need to know AI was used, not be slowed down by it.Low Risk, High Complexity
Help users stay oriented, guide them through complexity without overwhelming them or breaking flow.High Risk, Low Complexity
Make it crystal clear where the AI output came from, so users can trust it without needing to hunt for answers.High Risk, High Complexity
Slow things down just enough to help users focus, this is where safety matters most, and the UX should reflect that.
We paired each zone with reusable UX mitigators:
Positive friction
E.g., Slow stops, step-gates, reminders and confirmations
Explainable AI
E.g., Confidence scores, citations, explanations, references
Cognitive support
E.g., Streamlined layouts, clear cues, minimal noise
Reflective UX
E.g., Nudges to pause, guided reviews, clear boundaries.
The framework was baked into system guidance and supported by reusable components. Teams used it to self-assess, apply the right mitigators, and know when to pull in UX review. It helped Epic stay ahead of a major risk area and scale responsibly.
Increasing Transparency: AI Citations
Clinicians kept asking: Where did this come from? and Can I trust it?
I co-led the design of a standardized citation system to answer that. It became a core pattern in our gen AI framework, used especially in summaries and documentation tools.
(More examples and details available in the future.)
Building Trust: Human-in-the-Loop
As AI took on more tasks, giving users control mattered more than ever. I worked across teams to design patterns that kept users in the loop, without breaking their workflow.
We focused on feedback, explainability, and clarity around who’s responsible for what.
(More examples and details available in the future.)
Case study is In progress
Reinforcing Safety: Automation Bias
Automation bias happens when users over-trust AI, even when it’s wrong. In healthcare, that can lead to real consequences.
I partnered with Epic’s ethical ML group and led a multi-month initiative to tackle this head-on.
I created a framework teams could actually use: the Risk × Complexity Matrix.
It helped teams answer: How cautious do we need to be here?
We broke features into four zones:
Low Risk, Low Complexity – Keep it fast. Let users know AI was used, but don’t slow them down.
Low Risk, High Complexity – Guide the user clearly without adding noise.
High Risk, Low Complexity – Surface the evidence. Show where the AI output came from.
High Risk, High Complexity – Add focus and friction. Support safe, confident decisions.
We paired each zone with reusable UX mitigators:
Positive friction
E.g., Slow stops, step-gates, reminders and confirmations
Explainable AI
E.g., Confidence scores, citations, explanations, references
Cognitive support
E.g., Streamlined layouts, clear cues, minimal noise
Reflective UX
E.g., Nudges to pause, guided reviews, clear boundaries.
The framework was baked into system guidance and supported by reusable components.
Teams used it to self-assess, apply the right mitigators, and know when to pull in UX review. It helped Epic stay ahead of a major risk area and scale responsibly.
Increasing Transparency: AI Citations
Clinicians kept asking: Where did this come from? and Can I trust it?
I co-led the design of a standardized citation system to answer that. It became a core pattern in our gen AI framework, used especially in summaries and documentation tools.
(More examples and details available in the future.)
Building Trust: Human-in-the-Loop
As AI took on more tasks, giving users control mattered more than ever. I worked across teams to design patterns that kept users in the loop, without breaking their workflow.
We focused on feedback, explainability, and clarity around who’s responsible for what.
(More examples and details available in the future.)
Case study is In progress
Solving Enterprise-Scale UX Challenges
These were three of the most complex and impactful problem spaces I led - where the right design call shaped adoption, trust, and safety.
Epic Case Study
Scaling an AI Design System Across 80+ Clinical Apps
Company
Role
Lead UX Designer
Work
UX strategy & design, design system, visioning
Year
2022 - 2025
