INCIRCLE REVIEW // Dashboard & App / Design and Strategy Leadership

InCircle Review:
Faster Diagnoses Through UX

Overview

Ten hours. That's how long a neurologist spent parsing a single patient's history before rendering an opinion. This was before the current AI wave. InCircle was building machine learning into clinical workflows when most companies were still figuring out chatbots. The technology could parse the data. The design couldn't present it. 

The unforgiving interface wasn't a UI issue. It was a care delivery issue.

I provided design strategy and creative direction, guiding the UX strategist and design team while contributing a significant portion of the final design. The pitch was efficiency, but mapping the workflow revealed a different story. Neurologists care about confidence. If the interface creates doubt, the physician slows down regardless of what the data says.

Time is money, but it's more about the human need.

Four principles followed, each solving a specific failure in the existing product. Reduce sprawl: structured hierarchy so clinicians march toward an opinion at their own pace. Make it easy for everyone: the roadmap included insurance reviewers and nurse coordinators, not just specialists.

Clarity as the Product

Information at your fingertips: key interactions at the point of need, not behind navigation detours. Make it accessible: an ADA- and WCAG-compliant icon system where color, shape, and label all carry the same signal.

The most impactful decision was the interactive patient timeline: the entire history rendered as a navigable chronology where neurologists could spot treatment gaps and see the AI's recommendation in the context of the full record. The confidence score meant more when you could see what informed it.

The dashboard was modular by default. I'd built the grid in Figma using autolayout, so during a client video call I demonstrated drag-and-drop by simply rearranging cards in the design file. The client saw the experience, not the tool. That simplicity built deep trust.

What This Project Reveals

Ten hours to ten minutes. Not because the AI got smarter. Because the interface finally got out of the way.