Designing for Clinician Workflows: What I Learned from Building Real Healthcare Tools

Designing for healthcare isn’t just about creating digital tools—it’s about respecting clinical time. Every extra click, redundant alert, or misplaced control can slow down care and increase frustration. Working with Curie AI’s pulmonology teams taught me how to build interfaces that fit naturally into clinicians’ existing routines, rather than disrupt them.

Lesson 1: Start with Observation, Not Assumptions

Early on, we shadowed physicians and respiratory therapists during morning triage rounds. What looked like a simple dashboard interaction from the outside was actually part of a complex multitasking flow: clinicians were reviewing charts, listening to patient notes, and updating EHR systems—all at once.

Instead of trying to “reimagine” their process, we focused on augmenting it. For example, we designed AI-driven alerts that surfaced only the most relevant patient anomalies, embedded within the workflow they already used. No new tabs. No context switching.

Lesson 2: Speed and Clarity Trump Novelty

Clinicians operate in a time-sensitive environment. Our goal was to make the product invisible when it needed to be—fast, predictable, and frictionless. Every second saved from searching, scrolling, or guessing adds up across a clinical day.

• Simplified navigation to two clicks or less for common tasks like reviewing overnight events.
• Built keyboard-first interactions for faster input and chart updates.
• Standardized data visualizations across all patient modules so users didn’t have to re-learn visual conventions.

The best compliment we received wasn’t about beauty—it was about flow. “It feels like part of my shift, not another system,” one clinician said.

Lesson 3: Collaboration is the Real Design Tool

Great clinical UX is the product of continuous dialogue. We didn’t just hand over mockups; we co-created with doctors, nurses, and data scientists. We held feedback sessions during live pilot programs, letting clinicians annotate the UI in real time as they worked.

This process revealed insights no usability test could—like how physicians preferred to see patient trends summarized weekly instead of daily, or how a red alert color could increase anxiety in certain patient populations. We adjusted accordingly.

Lesson 4: Build for Trust, Not Just Usability

In medicine, design isn’t neutral—it carries responsibility. When we visualized AI-driven respiratory scores, we paired every output with a clear explanation of how it was derived. Transparency built confidence, which led to adoption.

Over time, clinicians came to rely on these insights as part of their daily rhythm. The design had quietly achieved what every tool should:help people do their best work without thinking about the tool itself.

Key Takeaway

Designing for clinicians is about humility. It’s not about reinventing medicine—it’s about amplifying expertise through clarity. Real progress happens when technology steps out of the way and lets clinicians focus on care.