Smart Triage
Designing AI-Powered Triage for Faster Diagnosis and Treatment
Quin | Amsterdam Netherlands 2024-2025
About Quin
Quin develops digital tools to make healthcare more accessible by bridging the gap between general practitioners, specialists, and patients. By simplifying care pathways and streamlining workflows, Quin helps overburdened healthcare providers manage growing patient demands with smart, technology-driven solutions.
Business challenge
Irritable Bowel Syndrome (IBS) affects 5.8% of Dutch adults, leading to long waiting times, increased healthcare visits, and a heavy burden on specialists. Many patients cycle through GPs, physical therapists, and alternative care before finally receiving specialized treatment. This inefficiency drives up costs—29% in primary care and 116% in secondary care after diagnosis.
How might we leverage AI to optimize gastroenterology triage while ensuring transparency, building trust with specialists, and maintaining high-quality patient care?
My role
My main role was designing interactive UI prototypes that guided implementation and secured hospital buy-in. While the UX designer led flow mapping, I co-developed wireframes with the UX and Service Designer. I also worked closely with developers to ensure the designs were practical, aligned with the design system, and ready for clinical use.
Approach and insights
To tackle this challenge, we first immersed ourselves in specialists’ workflows—pinpointing pain points, inefficiencies, and opportunities for AI-driven triage. Our goal was to design a solution that seamlessly integrates into existing processes while maintaining trust and usability. This process involved:
Researching workflow challenges with MDL specialists to improve triage and referrals.
Mapping user flows to align with specialist workflows and pain points.
Analyzing existing solutions to find gaps and opportunities.
Exploring explainable AI to ensure transparency and trust.
Wireframing and prototyping, refining through feedback.
Creating high-fi prototype for testing, implementation, and hospital pitches.
Collaborating with developers to align UI designs with development and the design system.
Continuous user testing to refine the product for real-world use.
Insights
We uncovered several key challenges in the gastroenterology triage process:
Time-consuming patient referrals: Specialists spend significant time reviewing patient history, referrals, and questionnaire responses before making decisions.
Many IBS patients don’t require a physical appointment: A large portion of cases could be managed digitally with structured guidance, reducing unnecessary hospital visits.
Long waiting times for simple advice: Patients often wait over a year, only to receive recommendations that could have been provided much earlier through a digital pathway.
Specialists’ time is misallocated: Gastroenterologists are overwhelmed with low-priority cases, diverting time from patients who truly need specialized care.
These insights shaped our design decisions, ensuring that our solution streamlined specialist workflows, reduced administrative burdens, and helped patients receive timely, appropriate care.
The Solution: Quin’s Smart Triage Tool
An AI-powered system that speeds up IBS diagnosis, reduces hospital visits, and supports specialists with transparent, explainable decision-making—while keeping them in control. This solution is part of Quin’s suite of products, designed to align with business goals by optimizing healthcare workflows and improving patient access to care. Key features include:
Step 1: Structured data access: Prioritizes urgent cases, automates assessments, and provides a clear clinical overview.
Step 2: Explainable AI analysis: Breaks down diagnoses with step-by-step logic, linked sources, and contributing factors.
Step 3: Personalized treatment selection & advice: AI suggests treatments, specialists refine, and patients receive clear, editable guidance.
High level user flow
The specialist moves through the triage process by scanning cases, reviewing summaries, checking AI suggestions, and making adjustments when needed. Each step supports clinical decisions while saving time and keeping control in the right hands.
Step 1:
Structured data access
Prioritizes urgent cases in a work list, automates assessments, and provides a clear clinical overview.
Cutting through data overload
In early research sessions with specialists, we found that one of the biggest challenges was scrolling through overwhelming amounts of patient data. Specialists needed structured, high-priority information at a glance, while retaining access to full medical records when necessary.
Solution
A Triage List prioritizing urgent cases for faster intervention.
A structured data panel, summarizing clinical information while allowing deep dives to raw data when needed.
AI-powered analysis that evaluates referrals and questionnaire responses, streamlining patient classification.
Step 2:
Explainable AI analysis
Displays AI-driven analysis of alarm symptoms, risk factors, and probable diagnoses based on clinical data. Provides a step-by-step breakdown with linked sources and contributing factors for transparency and trust.
Building trust through transparency
One of the most critical insights from our research was that specialists won’t trust AI unless they understand how it reaches decisions. They need visibility into the reasoning behind each recommendation from the raw information–in one screen.
Solution
A dedicated AI analysis panel that breaks down symptoms, risk factors, and diagnoses in an explainable way.
Expandable sections to reveal AI logic step by step, allowing specialists to verify and cross-check AI-generated insights.
Linked sources and contributing factors, reinforcing trust in AI recommendations.
AI as a support, not a substitute
Through iterative testing, it became clear that clinicians will not adopt AI-driven solutions unless they retain control over final decisions. They needed an intuitive way to override or adjust AI-generated outcomes.
Solution
A flexible editing mode where specialists can adjust symptoms, risk factors, and diagnoses.
AI-generated recommendations remain editable, ensuring specialists can tailor them to specific cases.
Positioning AI as a support tool, assisting triage rather than making autonomous decisions.
Step 3:
Personalized treatment selection & advice
AI suggests treatments, specialists select and refine treatment and advice, and patients receive clear, pensonalised guidance.
Making treatment personal, efficiently
Specialists want to provide personalized care, but they lack the time to manually draft advice for every patient. At the same time, patients are more likely to follow treatment recommendations when they feel tailored and personal.
Solution
AI suggests treatment options based on confirmed diagnoses in previous steps.
Specialists review and select the most appropriate plan.
AI generates clear, patient-friendly medical advice, editable by the specialist before being sent.
Outcomes
The AI-powered triage solution is anticipated to:
Accelerate patient diagnosis and treatment, significantly reducing wait times and improving access to care.
Reduce the workload of MDL specialists, enabling them to focus on complex cases while routine assessments are automated.
Enhance diagnostic consistency and accuracy, as AI-powered analysis provides standardized assessments based on comprehensive patient data.
Optimize workflows, lowering administrative burdens, reducing manual tasks, and improving specialist efficiency.
Improve patient experience, offering personalized digital treatment plans that minimize unnecessary hospital visits and enhance overall satisfaction with care.
As the solution is classified as a medical device, it must meet strict compliance regulations before deployment in clinical settings. The product is now prepared to undergo a clinical study to rigorously validate its effectiveness, ensuring it meets both safety and regulatory standards before broader adoption.
Final reflection
One of the biggest challenges in AI-driven healthcare is earning clinicians’ trust. Specialists don’t just need a faster solution; they need one that is reliable and makes sense. This project wasn’t just about building smart triage; it was about designing a system that truly supports specialists while respecting the complexity of their work. I learned that sometimes the best solutions deviate from conventions and are not the most obvious. Getting there meant questioning assumptions, letting go of ideas we thought would work, and staying open to change. However, the real key is working together, especially with the people who use it.
The people
Indeed, innovation in healthcare isn’t just about technology. It’s about people. The PMs kept us aligned on strategy and compliance, helping us balance ambition with regulations and overall business objectives. The UX designer’s ideas pushed innovation further, challenging us to think beyond the obvious. The Service Designer kept us grounded, asking critical questions and ensuring close collaboration with MDL specialists so our solution fit real-world workflows. The MDL specialists played a crucial role, continuously sharing their pains and experiences, helping us design a tool that truly met their needs. I, as the Product Designer made sure our designs came to life, aligning with user needs and business goals while facilitating smooth collaboration with developers and ensuring our work contributed to the growth and stability of the design system. And in the end, it was the Developers who made it all work, turning our ideas into a functional, real-world solution.
