AI-Accelerated Clinical Trial Finder

I worked on the UX strategy and design for a proof-of-concept clinical trials finder for Johns Hopkins All Children's Hospital. In just six weeks, our team needed to understand the problem, gather input from clinical stakeholders, understand families' needs, establish a visual and content direction, and turn it into a working prototype.

Clinical Trial detail page
Clinical Trial detail page
Clinical Trial Finder Home page presented on a laptop

Overview

The goal was to make it easier for families across Florida to discover relevant clinical trials, understand their options, and connect with the right support.

Working within a highly condensed timeline, I helped take the experience from understanding the problem and stakeholder needs through user insights, experience design, visual direction, and a working prototype.

Main Role: User Experience Designer, Strategist
Additional Role: AI-Assisted Developer
Time Frame: 6 weeks
Tools: Claude Code, VS Code, Figma Confluence

Main Role: User Experience Designer, Strategist
Additional Role: AI-Assisted Developer
Time Frame: 6 weeks
Tools: Claude Code, VS Code, Figma Confluence

Understanding the problem

Finding a pediatric cancer clinical trial can be overwhelming. Families often have to navigate ClinicalTrials.gov, interpret complex medical terminology, and contact hospitals individually to understand what options may be available.

We worked from real patient scenarios and community feedback provided by the client stakeholder team to understand what families needed to know:

  • What trials are available?

  • Where are they located?

  • Is this trial relevant to my child?

  • What does the trial actually involve?

  • Who can I talk to?

Finding a pediatric cancer clinical trial can be overwhelming. Families often have to navigate ClinicalTrials.gov, interpret complex medical terminology, and contact hospitals individually to understand what options may be available.

We worked from real patient scenarios and community feedback provided by the client stakeholder team to understand what families needed to know:

  • What trials are available?

  • Where are they located?

  • Is this trial relevant to my child?

  • What does the trial actually involve?

  • Who can I talk to?

Designing the experience

Find

Help families discover relevant trials through diagnosis, age, and location.

Understand

Translate complex clinical trial information into approachable, digestable content.

Connect

Give families a path to a human Nurse Navigator for additional guidance.

Building at speed

With only six weeks to move from discovery to a working proof of concept, we used AI as part of the design and development process. AI tools helped us rapidly explore ideas, generate and iterate on the prototype, establish reusable patterns, and accelerate development.

We also explored how AI could become part of the product itself — using Johns Hopkins' internal Claude API to generate plain-language summaries and translations of clinical trial information.

Stakeholder feedback, clinical scenarios, UX considerations, and QA continued to drive what made it into the experience.

With only six weeks to move from discovery to a working proof of concept, we used AI as part of the design and development process. AI tools helped us rapidly explore ideas, generate and iterate on the prototype, establish reusable patterns, and accelerate development.

We also explored how AI could become part of the product itself — using Johns Hopkins' internal Claude API to generate plain-language summaries and translations of clinical trial information.

Stakeholder feedback, clinical scenarios, UX considerations, and QA continued to drive what made it into the experience.

Wireframing a solution

I worked through the core experiences through low-fidelity, rapid wireframing using Figma's AI agent, focusing on how information could be organized and progressively revealed without overwhelming families.

The trial detail experience became an important part of this work. Rather than presenting clinical trial information as it appears in a database, we explored how it could be translated into clearer, more approachable information.

Additionally, we designed a direct connection to a Nurse Navigator so families always had a path to human support.

I worked through the core experiences through low-fidelity, rapid wireframing using Figma's AI agent, focusing on how information could be organized and progressively revealed without overwhelming families.

The trial detail experience became an important part of this work. Rather than presenting clinical trial information as it appears in a database, we explored how it could be translated into clearer, more approachable information.

Additionally, we designed a direct connection to a Nurse Navigator so families always had a path to human support.

Building a familiar experience

The portal needed to feel connected to Johns Hopkins All Children’s Hospital while establishing a visual language appropriate for a family-facing experience.

I helped shape the visual direction and reusable design patterns, creating a system that could carry consistently across the experience while remaining distinct from the broader Johns Hopkins brand.

The portal needed to feel connected to Johns Hopkins All Children’s Hospital while establishing a visual language appropriate for a family-facing experience.

I helped shape the visual direction and reusable design patterns, creating a system that could carry consistently across the experience while remaining distinct from the broader Johns Hopkins brand.

From design to prototype

The six-week timeline meant we had to move quickly from ideas to something tangible.

We used AI tools heavily throughout the process to accelerate exploration, iteration, and development. This allowed us to quickly turn design decisions into a working prototype and spend more of the limited timeline reviewing and refining the experience with stakeholders.

The prototype connected to ClinicalTrials.gov data, interactive maps, and AI-generated plain-language trial summaries and translations.

The six-week timeline meant we had to move quickly from ideas to something tangible.

We used AI tools heavily throughout the process to accelerate exploration, iteration, and development. This allowed us to quickly turn design decisions into a working prototype and spend more of the limited timeline reviewing and refining the experience with stakeholders.

The prototype connected to ClinicalTrials.gov data, interactive maps, and AI-generated plain-language trial summaries and translations.

Home page

Clinical Trial detail page

Resources page

Nurse Navigator intake

Guided clinical trial finder experience

Outcomes and Takeaways

In six weeks, we took the project from an early concept to a working proof of concept that demonstrated how a statewide pediatric cancer clinical trials experience could work.

The POC included:

  • Live Florida pediatric oncology trial data

  • Guided trial discovery

  • Interactive maps

  • Plain-language trial summaries

  • Multilingual content

  • Nurse Navigator intake

  • A scalable design and technical foundation

More importantly, the project gave Johns Hopkins Pediatric Cancer team a tangible experience to evaluate and a foundation for moving toward an MVP.

In six weeks, we took the project from an early concept to a working proof of concept that demonstrated how a statewide pediatric cancer clinical trials experience could work.

The POC included:

  • Live Florida pediatric oncology trial data

  • Guided trial discovery

  • Interactive maps

  • Plain-language trial summaries

  • Multilingual content

  • Nurse Navigator intake

  • A scalable design and technical foundation

More importantly, the project gave Johns Hopkins Pediatric Cancer team a tangible experience to evaluate and a foundation for moving toward an MVP.

Device mockups (laptop, tablet, mobile) displaying newly designed pages across clinical trial finder site

Jordan Nguyen

Let’s talk