In 2009, Fast Company profiled David Butler, Coca-Cola’s then vice president of design, who faced a fragmented brand. Across the globe, Coca-Cola’s visual identity had drifted as local teams and vendors improvised. Butler’s answer wasn’t a rigid rulebook but a new mindset: treat design as infrastructure.

He reduced Coca-Cola to its essentials: the red disc, the wave, the script, and reassembled them into a flexible system. Less a style guide and more a Lego set, it allowed employees to produce and vendors to build their own signage while keeping brand coherence. The goal wasn’t control but consistency through shared tools.

This Lego-like approach extended to Coke’s Freestyle machines, which let consumers mix their own drinks. The company no longer dictated flavor releases; it built a platform for others to design experiences. Design shifted from surface decoration to structural system. From something made to something that makes.

The analogy in software is even stronger. Infrastructure here means modular systems: tokens, components, interaction patterns that can be endlessly recombined. A design system becomes generative.

Add AI, and agents become the worker bees. They handle the construction tasks once done manually: coding interfaces, generating content, linking workflows. Infrastructure sets the grammar; agents compose with it. Teams can sketch intent, and agents assemble details while staying coherent because the system guides them.

The payoff is scale with coherence. Just as Butler enabled thousands of Coca-Cola touchpoints to remain on brand without micromanagement, software infrastructure plus AI agents could let organizations move at exponential speed without chaos. Infrastructure sets the rules of play; agents bring the energy.

This reframes design not as artifact-making but as laying tracks. Once the tracks exist, trains run, freight moves, commerce flows. Designers ensure coherence and resilience. Agents keep it in motion. Together they form a living infrastructure — one that builds not just things, but the capacity to keep building.

Design as Infrastructure: Building Systems That Build Themselves was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

Tech debt are the unwanted soot sprites of every software stack, lingering in dark corners, multiplying silently, and demanding attention at the worst possible moments — and, I might add, the bane of any designer hoping to create an imagined future state.

But with AI, we’ve stumbled on a curious opportunity: not just to clean house, but to install a self-dusting system. AI can help resolve long-standing tech debt by automating refactors, exposing hidden dependencies, and translating brittle legacy code into modular frameworks. For teams who’ve been kicking the can for years, that’s nothing short of a gift.

But before we declare victory, we have to ask: what kind of debt is AI creating in return?

This next-gen debt is subtle, even invisible. It lives in biased training data, aging foundation models, and black-box logic that resists inspection. It doesn’t throw 500 errors, it quietly distorts user experiences, misguides teams, and erodes trust. Are we ready for tech debt that isn’t just technical, but ethical and systemic?

As AI embeds deeper into our systems, tech debt becomes “meta.” It hides in the models we inherit, the prompts we forget to update, the assumptions we never question. So how do we maintain system health when the system is reasoning for itself? Is better documentation enough? Stricter reviews?

Probably not.

We’ll need AI that’s reflexive and antifragile: able to audit its own logic, recognize when its priors are stale, and flag when it’s building on shaky ground. Guardrails can’t just be static, they must adapt, evolve, even grow stronger under pressure. Can we design AI to carry its own quality assurance process in its psyche?

Because if we don’t, we’re just trading one kind of chaos for another.

Yes, AI might finally help us address the mess we’ve made. But we can’t afford to outsource our vigilance. The next wave of tech debt won’t look like spaghetti code. It’ll look like faulty cognition. And the only way forward is to embed a conscience in the machine.

AI can fix your tech debt, but what’ll fix AI’s? was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

Some of the best movies stick with us because we get to witness a real character arc: someone changes, grows, figures things out. That same kind of arc is what I think about when I reflect on growing up without and then with computers.

I was born before they were a thing. My early years were screen-free. That meant boredom, unstructured time, and figuring out how to entertain myself. It wasn’t always easy, but it built creativity, resourcefulness, and patience. There was no back button, just trial and error.

Computers arrived slowly. In my teens, they were more novelty than necessity. In college, they became useful. In grad school, the internet showed up in earnest. It was clunky and slow, but it changed everything. Suddenly, research, collaboration, even career possibilities looked different. I remember feeling the shift. This wasn’t just a new tool, it was a new context.

Then came the iPhone. For those of us already in the workforce, it landed as a productivity boost, not a social fix. We’d had just enough analog in our lives to use tech with intention, not impulse.

Now, we’re watching AI enter the picture, and it promises to be the biggest shift yet. I feel lucky to be at a stage where I have enough experience to put it in perspective, but still enough curiosity to dive in. I get to understand it not just from the outside, but from inside the work.

Not everyone gets to grow up with nothing and still catch the front wave of everything. Early Gen X is a weird, wonderful sweet spot, and I’m grateful to have a front-row seat.

The last unplugged generation was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

In product development, we often talk about product, design, and engineering as a balanced triad, each contributing their unique expertise, while influencing both partners. While ideally balanced, they don’t quite tackle the same kinds of problems.

One way to look at the difference is by considering how open-ended their work is.

Product tends to operate in the widest space, deciding what to pursue and why, based on fuzzy signals from the market, user behavior, and business goals. Design narrows the field, working to define how we might solve those problems in ways that are intuitive, elegant, and usable. Engineering typically works in the most constrained domain: how do we build this well? — a tough challenge, but often with clearer parameters.

This isn’t a value judgment. It’s just the nature of the work.

Now enter AI.

It’s not surprising that AI is showing early strength in engineering. It’s well-suited to technical, structured tasks: code generation, testing, optimization. These jobs live in the cleanest, most bounded problem spaces.

But when those are handled by machines, engineers get more room to think bigger. More integration and architecture. More systems-level work. More early-stage shaping. And, increasingly, a toe dipped into the design side.

Zoom out again: as AI begins to assist designers with pattern recognition, visual exploration, and user-interface generation, designers may find themselves pulled further into strategic territory — service design, product thinking, market fit.

What we’re seeing isn’t job loss, rather it’s role expansion.

As AI moves up the stack, it relieves pressure at the base. That gives humans space to climb. Engineers become more design-aware. Designers ask more business-shaped questions. Product managers think more deeply about the entire experience, not just the roadmap.

There’s a well-worn saying that AI won’t eliminate jobs — it’ll elevate them. This may be what that looks like for the product-design-engineering triad.

The work doesn’t go away. It just gets more interesting.

How AI may reshape the product-design-engineering triad was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

In The Fountainhead, Ayn Rand introduces the idea of “second-handers”: individuals who derive their direction not from their own judgment, but from the approval or actions of others. While her critique focused on people, the idea finds a quiet echo in today’s business world.

In highly competitive markets, it’s natural, almost reflexive, for companies to track what others are doing. But when competitive awareness becomes the primary compass, organizations risk losing touch with what matters most: their customers. The question subtly shifts from “What problems are we uniquely positioned to solve?” to “What are our rivals doing? And, how fast can we follow?”

Consider the well-documented contrast between Microsoft and Apple in the early 2000s. Microsoft often responded to Apple’s moves — sometimes directly, sometimes through positioning campaigns. Meanwhile, Apple appeared to be less reactive, focusing instead on delivering tightly integrated products that addressed latent user needs. It wasn’t just about speed to market — it was about clarity of purpose.

This isn’t to say competitor-awareness is bad. It’s essential. But if it overshadows customer-centric vision, it can lead to derivative work and missed opportunities for meaningful innovation.

As we enter a new era shaped by artificial intelligence, this lesson becomes even more critical. Success won’t come from being the first to ship an AI feature, but from using AI to solve real, often unglamorous customer problems in elegant, effective ways. The race isn’t to be first — it’s to be right.

The takeaway? Inspiration can come from many places, but conviction should come from within. Build with your customers in mind, not your competitors in sight.

The cost of focusing on your competition was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

Leveling up isn’t just about skill — it’s about perspective and curiosity

The NBA Finals are underway, and as always, I’m watching closely — not just for the highlight reels, but the moments between them: the adjustments, the mental grit, the players who rise when it matters most.

It brought me back to my own years playing basketball — not at an elite level, just as an average college player — but enough to see what separates those who rise from those who plateau. And it struck me — that arc mirrors what we face in product design.

In youth basketball, it’s all about fundamentals — dribbling, passing, rebounding, and shooting. In design, it’s visual foundations and user interfaces: grids, alignment, hierarchy, components. You build control through repetition and intent.

As you move up, the game gets more complex. You read the floor, play within a system, and make decisions in motion. For designers, this is the shift from crafting screens to thinking in systems and solving problems. You collaborate, absorb feedback, and balance user needs with business realities.

At the college level, it’s no longer just about individual skill — it’s how you elevate the people around you. On the court, that means leadership, anticipation, and adaptability. In design, it means taking ownership, driving clarity through ambiguity, and thinking systematically. Your impact is measured less by execution, and more by your ability to align teams and shape outcomes.

I never made it to the next level, but I saw glimpses of what it took. The players who advanced weren’t just talented — they adapted, stayed coachable, and never stopped learning. The same is true for the best designers and leaders I know. They’re sharp — but more importantly, they’re curious. Open. Always scanning the floor, always evolving.

Wherever you are in your journey, the next level will demand something new: sharper skills, better communication, broader thinking, deeper resilience. But always — always — it will require perspective. And curiosity.

Leveling up isn’t just about skill — it’s about perspective and curiosity was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

It’s easy to get swept up in the hype. AI can generate, optimize, predict, personalize — sometimes faster than we can even explain what we’re trying to do. But in all this momentum, have we paused to ask: Are we solving the right problem?

AI is impressive, but it’s not magic. It’s a tool — a powerful one — but still just a tool. It doesn’t define strategy, understand your customers, or know what success looks like. That’s still on us.

With AI in the mix, there’s a subtle trap: we start with the capabilities, not the needs. “What might we do with this model?” quickly turns into “What should we do with it?” — and sometimes skips over “What problem are we actually trying to solve?”

Let’s be honest, it’s tempting to let the shiny object take the wheel. But if we chase the tech without a clear destination, we risk building clever things that don’t matter. Or worse, things that create noise instead of value.

The fundamentals haven’t changed. We still need to be rigorous in identifying real user pain points. We still need to validate whether our ideas are meaningful, usable, and valuable. AI can help us do that faster — but it won’t do it for us.

So ask yourself and your team:

Are we letting the tool dictate the direction?

Have we gotten clearer on the problem — or just fancier with the solution?

If AI weren’t available, would this still be worth building?

The promise of AI is real. But it’s only as good as the clarity behind it. Strategy still matters. Critical thinking still matters… at least for now :)

AI doesn’t replace insight. It rewards it. was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

In 2019, AI was already transforming industries, but generative AI and large language models were still emerging concepts. Today, those technologies are reshaping the way we work, create, and interact. This series of articles from that time explores AI’s potential and challenges — discussions that remain just as relevant, if not more so, in the era of GenAI.

In AI we trust?

How might we design intelligent systems that inspire trust?

We’re presently experiencing a major shift in technology led by developments in artificial intelligence (AI). While we place our trust in technology every day, with AI, we’re starting to receive recommendations by machines. As a result, our perceived risks, whether real or imagined, raise new questions that must be answered before we can trust these systems: Does the system value the same things that I do? How does it make decisions? Am I safe?

Read the full article.

Design for trust: principle #1

User expectations and perceived risks drive trust requirements.

Here, we dive deeper into the first principle, grounded in empathy for our user with respect to how their mental models manifest as expectations and perceived risks.

Read the full article.

Design for Trust: Principle #2

Trust is a dynamic relationship. It is tentatively granted, then tested over time.

Here, we dive deeper into the second principle that addresses the dynamic and tenuous nature of trust during the customer journey.

Read the full article.

Design for Trust: Principle #3

When building trust, motivation is more powerful than demonstration or explanation.

Here, we dive deeper into the third principle that explores visceral trust, how it is ingrained and how it can make interaction more smooth.

Read the full article.

AI then and now: how past insights resonate in the age of GenAI was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

In 2019, I had the privilege of working with Alicia Staley, patient advocate and three-time cancer survivor, to develop a workshop that would eventually evolve into a Patient Insights Board where she and our design team engage with patient advocates to gain insights on important patient-centered topics. Through these workshops, we’ve been able to appreciate the challenges that patients face with their healthcare journey and specifically when joining clinical trials.

Our childhood memories of jabs, pills and procedures inform our visceral response to doctor’s visits. In the best case, we‘re going for a preventative checkup. In other cases, there’s a problem — and the visit can be stressful. Beyond medical visits, clinical trials present a lifeline when standards-of-care have proven inadequate. Some of us, like Alicia, have been through the process, while many of us know someone who’s reached this crossroad. Sometimes the outcome of a trial may not benefit them directly, yet they’re driven by the desire to help others who may follow. We hear stories and witness bravery, optimism, resilience—heroism.

The 10 principles of patient-centric design are for these heroes. The principles have been developed over five years of interactions with patients, patient advocates, healthcare professionals, and clinical trial administrators. They’re meant to guide contributors to the patient experience, to be the calm in the chaos of each patient’s healthcare journey. We hope you find them useful.

Principle #1 Build trust

Trust is a dynamic relationship. It is tentatively granted, then tested over time. Manage patients’ expectations and perceived risks throughout their experience.

Transparency is often referenced as a key ingredient to trust building and it’s no different here. Be open and honest about personal information and data sharing, potential conflicts of interest, and the risks and benefits of participation. Educate patients of anticipated burden, both financially and with respect to time commitment.

Build trust by opening channels for feedback and communication. This demonstrates a willingness and capacity to listen, learn and support.

For a design deep dive, see how we create trust empathy maps to identify trust-building opportunities.

#2 Develop a good bedside manner

Put patients at the heart of everything you do. Make the patient feel cared for, comforted and human before, during, and after their treatment or trial.

Start with a step-by-step agenda and provide the patient with a sense of place and time. This may seem elementary, but there are many examples where the patient is left to wonder “what’s next?” or “how much longer?”

Inevitably, despite every effort to guide and nurture the patient through an experience, some will inevitably lose their way. In these cases, show the human on the other side by providing convenient help and support.

#3 Be inclusive

Good design is accessible to all.

Patient diversity is essential in developing meaningful outcomes in clinical and medical research because it ensures that findings are applicable across abilities, backgrounds, medical conditions and cultures, leading to more accurate and effective healthcare interventions.

#4 Empower with knowledge

Keep the patient informed and educated about their options so they can make good decisions.

Patients should be prepared to participate. This means providing educational resources during the decision making process and at touch points throughout the process. And, don’t assume that the patient understands. Always confirm their understanding.

When the process or procedure is complete, be sure to share data back to the patient in a format that’s easy to understand, draw insights from, and allows the patient to share data with selected family members, their primary care provider and healthcare teams.

#5 Give the patient control

Allow the patient to move at their own pace. Encourage interaction and dialogue, including feedback.

Empower patients to control their healthcare journey by providing them with flexibility to participate on their terms, choosing where, when and how they engage. Allow adequate time for patients to ask questions and to make informed, shared decisions with their loved ones and healthcare team.

#6 Celebrate wins

Acknowledge patient achievements and motivate their next steps.

When a process is completed, don’t just move on to the next task. Take the time to recognize each milestone in the process — high five! In gaming terminology, that’s called celebrating “win states.”

#7 Don’t make them think

Every element of a design should be in service of the intended patient action or outcome.

Reduce cognitive load whenever possible. While you may think a designed interaction makes perfectly good sense, it’s but one of many tasks in a patient’s busy day. They’re not reading, they’re scanning during a spare moment. Pursue simplicity relentlessly.

#8 Don’t reinvent the wheel

When possible, use standard patterns and conventions.

Utilize familiar design patterns to tap into patients’ pre-existing mental models, offering a sense of comfort and ease of use. There are places to be creative, but navigation and typical computer interactions are not two of them.

On the other hand, intuitive patterns go beyond familiarity by anticipating patients’ needs and behaviors, often employing innovative or unconventional design solutions that feel natural and effortless.

Pay attention to when familiar will suffice and when non-standard yet intuitive designs are warranted.

#9 Choose words carefully

Messaging and instructional content are critical to the patient experience.

Writing is often overlooked in the design process, yet it’s one of the most critical. It’s the aspect of our designs that patient advocates have commented on most often.

Speak in natural, non-technical language. Find the balance between casual, friendly language and a tone that respects context.

And, consider how certain words may fall upon patients’ ears. Historically, participants in clinical trials have been referred to as “subjects.” This terminology still exists and fails to acknowledge the patient’s voluntary role as a participant and partner in the process.

Am I data, or am I going to be seen as a real person?*

When I hear the word subject, I think of a criminal.*

I feel like one in a million, not one in a million.*

I think of an inanimate object—something that isn’t a person.*

*Quotes from Medidata Patient Insights Board members

#10 Pay attention to the details

Proper typography and consistent design elements are critical to a trustworthy experience.

A consistently designed experience, from branding, to communication style, right down to font selection, all play a role in providing a trustworthy experience. Patients may enter the system from a variety of touchpoints (posts, email, notifications, etc.) and you want them to know they’ve come to the right place.

Beyond consistency, quality is essential. If you don’t bother to dot your i’s and cross your t’s literally, what does that say for your healthcare practices?

Final thoughts

Patient-centered design is a powerful approach to improving the patient healthcare journey. Trust building is the foundation on which all future interactions unfold. It’s built through transparency, listening and learning—by showing that there are humans on the other side. From here, patient empowerment is constructed through knowledge and control, and only through inclusivity can we ensure better outcomes for all.

We would love to hear your thoughts on how these principles resonate with your experiences, whether as a provider, designer or patient. Your feedback is invaluable in refining our approach and ensuring that future solutions truly meet patient needs.

Acknowledgements

Special thanks to Alicia Staley and the Patient Insights Board at Medidata for their time, dedication and inspiration; to Ben Berte and the Patient Cloud design team for their pursuit of better patient experiences, facilitation of patient-centered design workshops, and to Jesi Zhang for the design examples presented here. In memory of Linnea Olson and Sheila Johnson whose perspectives greatly influenced our understanding and development of these principals.

Ten Principles for Patient-centered Design was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.

The decision to participate in a clinical study often comes at an emotional and vulnerable time for patients when more standard treatments have proven ineffective. Patients may have questions about the process, they may even have preconceived notions, and these notions can contribute to uncertainty and fear.

With big pharma at the bottom of the industry-positivity rankings, recruiters face an uphill battle convincing patients to join a study. While there are many factors that contribute to a successful recruiting campaign, building and nurturing patient trust plays an important role in smoothing the inherent speed bumps along the way.

But how do we identify trust-building opportunities among a group of patients who may have very different goals and perspectives? Statistics can provide indicators of high churn along the recruitment funnel, but the causes of drop out can only be identified by talking directly with patients.

In this article, we share a design process that we use to understand patients’ goals, expectations and perceived risks, and ultimately to identify and prioritize trust-building opportunities.

Mapping Patient Trust

While trust itself is an emotional, social construct, our ability to build trust or more appropriately to avoid losing trust with patients, can be distilled to understanding their personal goals, expectations (particularly false expectations), and perceived risks.

Knowing a patient’s personal goals helps us to understand what motivates them to act. Goals also help to contextualize expectations and perceived risks.

False expectations are the result of the patient not understanding how a clinical study works or a predisposition based on previous experiences. For example, a patient may assume that frequent visits to the clinician’s office are required even though some studies may provide a more convenient means of participating from home.

Perceived risks, whether warranted or not, are equal in the mind of the patient. A patient considering a vaccine trial, for example, may fear becoming infected by the virus. For traditional vaccines, this risk may be warranted, while for newer vaccines that don’t incorporate the actual virus, the risk is perceived. In either case, the patient’s concern must be addressed if they are to move forward.

Trust-building opportunities arise at the intersection of false expectations and perceived risks. Consider the following trust-building priority space:

An Example

We facilitated a workshop with five patient advocates to understand trust factors that could influence a patient’s decision to join a clinical study. We asked the following questions:

1. What do you hope to get out of participation? How do you want to feel along the way?
2. What are your expectations and how do you think the process works?
3. What are your uncertainties, concerns and fears?

We then identified emergent themes and derived trust-building opportunities.

The results were summarized in a trust-empathy map:

The trust-building opportunities identified and prioritized were:

1. How might we add transparency to the study’s data sharing practices?
2. How might we clearly explain the potential health risks associated with participation?
3. How might we assure patients that the study has their best interests in mind?
4. How might we better manage patient expectations in terms of the personal attention and care they will receive?
5. How might we add transparency to the financial expenses that the patient may incur during and following the trial?
6. How might we help patients find support amongst their peer communities?

Summary

Participating in a clinical trial is a commitment with associated risks and benefits. In addition, patients may deliberate with perceived risks based on their understanding of the process and business of clinical trials.

In this article, we introduced a systematic approach to identifying trust-building opportunities with potential clinical-trial participants:

• Understand patients’ goals, expectations and risks.
• Derive and prioritize trust-building opportunities based on perceived risks and false expectations in the context of personal goals.

With a research-backed, prioritized list of trust-building opportunities, we can now move confidently to ideation with the goal of recruiting more educated and empowered patients who in turn will be more likely to make an informed decision. Those who decide to participate are more likely complete their study.

Building Patient Trust was originally published in 2ejk on Medium, where people are continuing the conversation by highlighting and responding to this story.