ParaDocs Relay
Role
Duration
Who they are?
ParaDocs is a healthcare technology company focused on improving clinical documentation workflows, especially for chronic care. Their solutions integrate AI-generated insights into electronic health records (EHRs), helping clinicians reduce cognitive load and spend more time on patient care.
Primary care physicians often manage patients with multiple chronic conditions, requiring them to sift through years of clinical notes during brief visits. This results in cognitive overload, increased documentation time, and a higher risk of missing critical information—especially when manually compiling the history of present illness (HPI). The challenge was to design a solution that could intelligently summarize relevant patient data and fit seamlessly into existing EHR workflows without disrupting the provider’s routine.
The Challenge
Working with a clinical partner
We kicked off the project with a Zoom meeting to align on the problem space and project goals. While I led the end-to-end design process—from wireframes to final UI—my collaborator provided valuable clinical insights that shaped the content and functionality. Regular Zoom check-ins helped us translate those insights into a solution that fit seamlessly into provider workflows.
01. Minimizing typing and clicks during HPI documentation
02. Trust and clarity in AI-assisted content
03. Seamless fit into existing EHR workflows
Key user
Who are the users
Primary users
→ Primary care physicians managing patients with chronic conditions like diabetes, hypertension, and heart failure
Internal medicine and family medicine providers
Project outcomes
→ 40% faster HPI documentation
Clinicians were able to complete the history of present illness section more quickly using AI-generated summaries.
→ Improved trust in assisted documentation
Providers reported higher confidence when reviewing and editing condition-specific summaries, leading to greater adoption.
→ Designed for EHR integration
The solution was structured to meet SMART on FHIR standards, making it suitable for integration into Epic workflows.
Key areas for usability focus
Project goals and expectations
Kickoff meeting summary
The design was guided by the expectation that any new workflow must align with how clinicians already document in EHRs—without adding complexity. The goal became clear: enable physicians to quickly review, trust, and edit AI-generated summaries within the HPI section, reducing documentation time while maintaining clinical accuracy.
AI-assisted interface for faster and smarter HPI documentation
Design decisions
Key interaction
What I learned
1. Inline dropdown over modal views
We chose an inline dropdown for summary selection to maintain clinical context and minimize cognitive switching. Modals were avoided to keep providers anchored within the HPI workflow.
2. Grouping summaries by condition
Summaries are categorized by chronic condition (e.g., diabetes, hypertension) to match the way clinicians mentally organize cases and reduce scanning time.
3. Editable text blocks
Each summary is inserted as editable text, giving providers full control to modify tone, phrasing, or medical nuance before finalizing the note.
4. EHR-aligned layout
The interface mirrors common HPI input fields seen in EHRs like Epic or Cerner, reducing the learning curve and making it plug-and-play for future integration.
The core value of ParaDocs Relay lies in how effortlessly clinicians can insert AI-generated summaries into the HPI section.
By simply selecting a relevant condition summary from the dropdown, providers bypass 8–10 lines of manual typing. This interaction reduces documentation time, limits cognitive effort, and keeps the clinician focused within the EHR flow—improving both speed and user satisfaction.
Designing for clinicians taught me that simplicity isn't just about clean UI—it's about respecting their time, mental load, and workflow. Working closely with a clinical partner helped me translate real-world pain points into design decisions that feel intuitive and valuable. I also learned how important it is to balance AI automation with provider control to build trust in healthcare tools.
fin.