Solace
Solace is a Series B healthcare startup focused on simplifying healthcare in the U.S. by having advocates guide patients through complex processes. As part of a three-day hiring assignment, I created an AI Summary Panel to help advocates extract information and generate next steps from call transcripts.
❋ Role
UX Designer
❋ Skills
Figma
❋ Tools Used
Product Design
AI Interface Design
❋ Timeline
3 days
Existing Transcript Design
Solace’s healthcare advocate dashboard had a transcript feature, which allowed the advocate to review phone call conversations with patients. But it required manually scanning through dense lines of dialogue to find critical details.
My challenge was to design an AI-powered summary panel that showed key insights, linked them directly to the transcript, and suggested next steps. But I not only had to design the panel, I also had to fit it into the existing system.
Process/Approach
To understand how an advocate would use a transcript, I wrote a fictitious conversation that came directly from personal experience dealing with healthcare challenges. This allowed me to simulate real dialogue, and only after writing this transcript could I put myself in the shoes of a Solace advocate who needed to quickly extract information.
I then sketched out some layout and interaction options. I explored what features the AI summary panel would offer, and I ended up going with these three features: a summary, key patient details, and next steps.
Implementation
The final design is an AI summary panel with three parts: a call summary, key details about the patient, and actionable next steps. Each item is numbered and linked to the corresponding line in the transcript, which helps advocates quickly reference the original dialogue.
This act of connecting the information in the AI Summary to the transcript was critical to include, because in healthcare, there can absolutely be no room for AI fabricating facts. I had to make it clear that this AI Summary just gives you a bullet-point version of the information that was said in the call. Its only task is concerned with making the advocate’s work quicker.
Reflection
This assignment strengthened my ability to design user experiences in a data sensitive environment. It wasn’t just about making a positive user experience, it was also about crucial information that had to be handled correctly. I liked that there was a stronger sense of responsibility here to reduce the risk of missed information, and to use AI to enhance productivity.