PXZEN
All case studies
Healthcare / HealthTech
Product Concept

AI Triage Assistant

AI that helps patients understand what happens next. A healthcare AI triage assistant designed to reduce uncertainty around waiting, route patients to the appropriate care pathway, and help clinical teams manage patient flow more effectively.

Product Strategy
UX/UI Design
AI Interaction Design
Healthcare UX
Conversational AI
Core outcome · case-study target, not measured data50% reduction in patient wait-time confusionInstead of leaving patients wondering “When will I be seen?”, the experience gives them clear guidance about what happens next.
02 — Client / Business context

Designed for care settings where patient flow never stops

This concept represents a modern healthcare provider, urgent-care network, clinic or hospital environment where patient volume fluctuates throughout the day and patients arrive with different symptoms, urgency levels and expectations.

Meanwhile, healthcare staff need to
  • Understand patient needs quickly
  • Route patients appropriately
  • Prioritize urgent situations
  • Reduce unnecessary administrative work
  • Keep patients informed
  • Manage unpredictable patient flow

The challenge wasn’t simply to build an AI chatbot

“What happens next, and when should I expect help?”

Waiting is not always the biggest problem. Uncertainty is.

Patients don’t just need a place in the queue. They need clarity about the journey.

03 — The problem

Patients may not know

Where they should go
What information they need to provide
How urgent their situation is
Why someone else was seen first
How long they may need to wait
Whether they should continue waiting
When they will receive an update
What happens after check-in
This creates additional pressure for staff
  • Repeated “How long will it take?” questions
  • Incomplete patient information
  • Manual intake processes
  • Poor visibility into patient needs
  • Staff interruptions
  • Confusing waiting-room communication
  • Inefficient routing
04 — Business goals

Goals across three connected experiences

Patient Experience

  • Reduce uncertainty
  • Improve communication
  • Make check-in easier
  • Explain next steps clearly
  • Reduce unnecessary waiting-room frustration

Operational Experience

  • Collect structured information earlier
  • Reduce repetitive administrative questions
  • Improve patient routing
  • Give staff better visibility
  • Support more efficient patient flow

AI Experience

  • Make AI assistance understandable
  • Keep humans in control
  • Escalate appropriately
  • Avoid presenting AI as a diagnosis system
  • Provide transparent communication
05 — User research

Research on both sides of the healthcare experience

Patient interviews explored
  • What causes anxiety during waiting?
  • What information do patients want?
  • What makes a healthcare interface trustworthy?
  • What do patients find confusing during check-in?
  • When do patients want to speak to a human?
Staff interviews explored
  • What information is repeatedly requested?
  • Which intake tasks consume staff time?
  • Where do patients get confused?
  • What information is useful before a patient is seen?
  • Where does patient flow break down?
Research insight

The waiting experience becomes significantly more stressful when patients don’t understand what is happening.

So the product strategy focused not only on reducing waiting, but on reducing waiting-time uncertainty.

06 — Personas

Four people the experience had to work for

Patient

Goal: Get help and understand what happens next.

Uncertainty
Long waits
Confusing instructions
Anxiety
Difficulty knowing when to ask for help

Caregiver / Family Member

Goal: Understand the patient's status and next steps.

Lack of information
Repeated questions
Difficulty navigating the facility

Front Desk Staff

Goal: Process patients efficiently.

Repetitive questions
Manual intake
Incomplete information
Interruptions

Clinical Staff

Goal: Receive useful structured information before interacting with the patient.

Inconsistent information
Time pressure
Information scattered across systems
07 — Patient journey

Arrive → Check In → AI Triage → Routing → Wait → Updates → Human Care

The AI experience is designed to reduce uncertainty while keeping clinical decisions under appropriate human oversight.

Before
1Arrive
2Check in
3Wait
4Ask staff
5Wait
6Ask again
7Eventually seen
After
1Arrive
2Digital check-in
3AI-guided questions
4Appropriate care pathway
5Clear waiting information
6Status updates
7Human clinical evaluation
08 — AI triage opportunity

The core of the product

The AI assistant guides the patient through a structured conversation, asking relevant follow-up questions to collect the information needed for routing.

“What brings you in today?”

The system helps determine the appropriate administrative or care pathway according to predefined clinical protocols and escalation rules.

Important safety principle

The AI should not independently diagnose the patient.

Instead, it can:

  • Collect information
  • Identify potential urgency signals for escalation
  • Route according to approved protocols
  • Explain next steps
  • Request human assistance when appropriate
09 — AI triage experience

Four steps, always with a human safety net

Step 01

Patient Check-In

The patient enters basic information: name, reason for visit, basic symptoms and relevant context.

Step 02

Conversational Questions

Instead of a long medical form, the AI asks focused questions conversationally — “Can you tell me what you're experiencing today?” then “When did this start?” — adapting as the patient responds.

Step 03

Safety Escalation

If responses suggest a potentially urgent situation, the assistant stops the normal waiting workflow: “Please speak with a healthcare professional immediately” — and guides the patient to the human escalation pathway.

Step 04

Routing

For non-emergency scenarios, the system suggests an appropriate pathway — Primary Care, Urgent Care, Nurse Review or Specialist Pathway — presented as routing, never a diagnosis.

10 — Wait-time clarity

The key differentiator: unknown waiting becomes visible progress

Instead of “Please wait,” the interface shows exactly where a patient stands.

Your current status
1Check-in complete
2Information received
3Waiting for clinical review
4Next update expected soon
What the patient sees

You're checked in.

Your information has been received.

A member of the care team will review your information.

We'll notify you when your next step is ready.

Where reliable operational data exists, the system may show an estimated wait range — never a more precise estimate than the underlying data can support. The principle: communicate confidence honestly.

11 — AI assistant & human-AI handoff

Available while waiting — and clear about its limits

Patients can ask
  • What happens next?
  • Have I completed check-in?
  • What should I do if my symptoms get worse?
  • Where should I go?
  • When will I receive my next update?
  • Can I update my information?

For clinical concerns, the assistant directs users toward human care rather than attempting to replace a healthcare professional.

AI handles
  • Basic intake
  • Information collection
  • Status communication
  • Administrative guidance
  • Routine questions
Human handles
  • Clinical assessment
  • Diagnosis
  • Treatment decisions
  • Complex cases
  • High-risk situations
  • Exceptions
  • Patient concerns requiring professional judgment

A clear AI → Human handoff, by design.

12 — Staff dashboard & AI-assisted summary

Better information before staff ever interact with the patient

The objective is not to replace staff decision-making. It's to give staff better information before they interact with the patient.

Dashboard queues
Patients Waiting
Needs Review
Escalated
In Progress
Completed
Each patient record includes
  • Check-in status
  • AI-collected information
  • Routing status
  • Time in queue
  • Escalation status
  • Staff notes
AI-assisted patient summary
AI-generated
Reason for visit
Persistent cough
Started
5 days ago
Reported symptoms
Cough, fatigue
AI status
Requires clinical review
Patient concern
Symptoms worsening
Action
Clinical assessment

Always clearly identifiable as AI-generated / AI-assisted information, reviewed according to the organization’s workflow.

13 — Information architecture

Patient experience, separated from operational experience

Patient side

Home

  • Check In
  • AI Triage
  • Status
  • Wait-Time Updates
  • Messages
  • Help
Staff side

Dashboard

  • Patient Queue
  • Patient Details
  • AI Summary
  • Escalations
  • Patient History
  • Communication
  • Analytics
14 — Wireframing

Low-fidelity, high-clarity

Patient Check-in
AI Triage Conversation
Escalation Screen
Care Pathway
Waiting Status
Wait-Time Update
AI Assistant
Staff Dashboard
Patient Summary
Queue Management
Accessibility
Readability
Clear hierarchy
Low cognitive load
Large touch targets
Simple language
Stress-aware interaction design
15 — Healthcare design system

Accessibility and clarity over visual complexity

Typography, color, spacing, buttons, forms, cards, alerts, status and progress indicators, AI message components, patient cards, queue components and dashboard components.

Spotlight — patient status states
Check-in Complete
In Review
Waiting
Needs Attention
Escalated
Completed
16 — Final UI / Core screens

The ten screens that carry the experience

01

Patient Welcome

Simple and reassuring entry point.

UX: Sets a calm tone before anything clinical happens.

02

Digital Check-in

Fast, accessible patient intake.

UX: Large touch targets, plain language, minimal steps.

03

AI Triage Conversation

Conversational information collection.

UX: Feels like a conversation, not a medical form.

04

Urgent Escalation

Clear human-care escalation pathway.

UX: Unmistakable, immediate, never buried in the flow.

05

Care Pathway

Explains what happens next.

UX: Answers “what now” in one glance.

06

Waiting Status

Shows patient progress.

UX: Replaces silence with visible movement.

07

AI Waiting Assistant

Answers administrative questions during the wait.

UX: Available without requiring a staff interruption.

08

Staff Dashboard

Shows patient flow and operational status.

UX: One view of who's waiting, reviewed or escalated.

09

AI Patient Summary

Concise structured information for staff review.

UX: Cuts intake reading time before a patient is seen.

10

Queue Analytics

Insights into patient volume, waiting time, routing, escalations and completion.

UX: Operational visibility, not clinical prediction.

17 — Before vs after

From unclear waiting to a guided, transparent journey

BeforeAfter
Unclear waitingVisible status
Repeated questionsAI-assisted information
Long intake formsConversational intake
Manual routingProtocol-guided digital routing
Information scatteredStructured patient summary
Patient uncertaintyClear next steps
Staff interruptionsSelf-service guidance
AI as chatbotAI as workflow assistant
No clear handoffHuman-AI escalation
18 — Measuring success

Defining “wait-time confusion” instead of treating it as a vague metric

Patient Questions

How frequently patients ask “When will I be seen?”

Status Understanding

Percentage of patients who correctly understand their current status.

Support Requests

Number of repetitive waiting-related questions directed to staff.

Task Completion

Percentage of patients completing check-in and triage without additional assistance.

Escalation Accuracy

Whether predefined escalation pathways are appropriately triggered and reviewed.

19 — Results / Impact

Primary outcome

Case-study target · not verified real-world data

50% less confusion around waiting.

The product reduces confusion around waiting — not necessarily the clinical waiting time itself. That distinction matters, and we’re not claiming otherwise.

Better patient communication
Lower waiting-room anxiety
Fewer repetitive status questions
Faster information collection
Better staff visibility
More structured patient intake
Clearer patient journeys
More efficient human-AI collaboration
20 — Business value

Why this matters to the people who'd actually run it

For Patients

Know what's happening without constantly asking.

For Healthcare Staff

Spend less time answering repetitive status questions and more time supporting patients.

For Healthcare Providers

Create a clearer, more scalable patient-flow experience.

For HealthTech Companies

Turn AI into a practical workflow rather than another chatbot.

21 — Future AI roadmap

Operational decision support, not autonomous clinical decisions

Phase 01

AI Intake

Collect structured patient information.

Phase 02

Intelligent Routing

Support protocol-based routing and escalation.

Phase 03

AI Waiting Assistant

Provide contextual status and administrative guidance.

Phase 04

Staff Copilot

Summarize patient-provided information and surface relevant workflow actions.

Phase 05

Predictive Patient Flow

Help teams anticipate patient volume, queue pressure, staffing needs and bottlenecks — as decision support, not autonomous clinical decisions.

22 — Human-centered AI principles

The product's AI principles

01

AI Assists, Humans Decide

AI supports the workflow but does not replace clinical judgment.

02

Safety Before Automation

Potentially urgent scenarios trigger appropriate human escalation.

03

Explain the Next Step

Patients should always understand what happens next.

04

Never Pretend to Know

The AI communicates uncertainty instead of confidently guessing.

05

Keep the Patient in Control

Users can request human assistance at any point.

06

Make AI Visible

Clearly communicate when the user is interacting with AI.

23 — Pxzen healthcare AI design framework

Understand → Simplify → Assist → Escalate → Communicate → Scale

Our signature approach to designing healthcare AI products.

Understand

Understand patients, staff and operational workflows.

Simplify

Turn complex healthcare processes into understandable interactions.

Assist

Use AI to reduce repetitive work and guide users.

Escalate

Know when the experience needs a human.

Communicate

Make status, uncertainty and next steps clear.

Scale

Create a flexible foundation for future healthcare AI capabilities.

This is a design concept, not a clinically validated product. The AI assists with intake, routing and communication — it does not diagnose, treat or make autonomous clinical decisions, and every escalation path routes to a qualified human.

Building an AI Product for Healthcare?

Pxzen designs human-centered AI experiences that turn complex healthcare workflows into clear, trustworthy and usable digital products.