Inspiration

Life's biggest changes come in waves, and a health scare is one of them. It's 2am and something feels wrong. You're alone, scared to overreact and scared to wait. You search your symptoms and get ten answers, each scarier than the last. You don't want a diagnosis, you want someone to guide you to find the next step.

For this challenge we reimagined Help Me Decide, Baylor Scott & White's care-navigation tool in MyBSWHealth, so people can find it easily, trust its guidance, and understand their recommended next step. We named our redesign Sage, after an armadillo guide who carries steadiness with you. The name stands for Support, Advocacy, Guidance, and Empathy, the four things our research showed people need when they're unsure of themselves.

What it does

Sage helps people decide whether a symptom needs the ER, urgent care, a nurse, a doctor's visit, or rest at home. It gives guidance to make a decision, not a diagnosis. It's a conversation, not a form. The chat appears as if you're texting an actual person for a consultation. It gives you a clear recommendation. The result states the level of care you need, a primary action, nearby care options, and a short "why" so people understand and can explain the reasoning. There's honest about uncertainty. Sage explains what it can and can't know so the recommendations feel transparent and safe. An emergency screen that looks like an emergency so people don't overlook it. This is built for different needs: we've designed a full set of Figma frames for audio-accessible English, English, Spanish, and high-contrast versions.

How we built it

First we started with research, then we designed accordingly. We sent out a survey to a multitude of our friends (those who are tech-savvy and who are not), parents, siblings, etc. Overall we had 31 respondents rating a screenshot of the original emergency recommendation screen from 1 to 10 for perceived urgency. They also answered open-ended questions about when they'd use it, what they want, how urgency could improve, how it compares to alternatives, and what features to add. Then we analyzed the data we collected. We categorized the written answers into four themes: decision support, trustworthy information, actionable next steps, and calm reassurance. We excluded filler responses to keep the data relevant. Next, we designed and built an editable prototype in Figma, with a white, black, and yellow palette and blue and purple highlights. Lastly, we used our research-driven decisions to create distinct indicators in our design. We used all the common themes to reformat the wording, the colors, the text hierarchy, and action statements to help people use the tool to its full potential.

Challenges we ran into

Making urgency visible without causing panic. The original screen was read inconsistently: 45.2% rated it high urgency (8–10), 29.0% moderate and 25.8% low, with a standard deviation of about 2.7. Nobody rated it a 10. Designing for people who are stressed or tired, which meant cutting cognitive load while still explaining recommendations enough to build trust. We researched literacy rates and found that most medical information is aimed to be at 6th-7th grade literacy rate to improve comprehension of data. Designing for blind and low-vision users. A screen-based chat doesn’t translate directly to audio, so we rethought how options are presented, with numbered “Press N:” choices that read cleanly aloud.

Accomplishments that we're proud of

A design process where every major decision traces back to what real people told us. A full accessibility set: English, Spanish and high-contrast frames, each with a distinct audio experience for blind users, not a simple add-on. A recommendation screen designed to be understood in seconds, even at 2 a.m. A cohesive brand: Sage the armadillo, four clear values, and a consistent visual system. A research story that states its limits honestly.

What we learned

One main thing we learned was people want confidence, not a diagnosis. Urgency is a design problem, not just a wording problem. The words of the page had said emergency while the interface looked calm, this allowed people to overlook it, this can't happen. Trust comes from explaining, not reiterating generic information. Accessibility improves the experience for everyone and it shouldn't be an add-on, it should be one of the leading factors of a design.

What's next for SAGE

Test the redesign with a larger group with a more varied age range to see whether the new emergency screen is read more consistently. Validate with clinicians so triage logic and recommendations are clinically sound. Test with blind and low-vision users and real screen readers to refine the audio flow. Expand to languages beyond English and Spanish. Measure behavior, not just perceived urgency, since self-reported reactions don't reflect what people would do in a real emergency.