Mobility Ambassador

Mobility Ambassador

Seattle’s government-backed app for convenient and inclusive public transit access.

AI | Conversational Design | Full-stack To C Project
An AI-powered transit assistant designed to provide accessible, real-time, and personalized support.

Launch Project by King County Metro

AI | Conversational Design |To C Project
An AI-powered transit assistant designed to provide accessible, real-time, and personalized support.

Timeline

2024.10-2025.3

Role

Product Designer

User Researcher
Visual Designer

Team

Me (PM and Designer)

2 Engineer

Team

Me (PM and Designer)

2 Engineer

Approach

Reflection

Impact Measured Through Iterations

Task Success Rate

45%

User Satisfaction Rate

10%

User Drop-Off Rate

35%

Impact Measured Through User Metrics

Task Success Rate

45%

User Satisfaction Rate

10%

User Drop-Off Rate

35%

Impact Measured Through User Metrics

Task Success Rate

45%

User Satisfaction Rate

10%

User Drop-Off Rate

35%

Understand Stakeholder

King County , as government department, aims to create a local travel helper ambassador——also supporting underserved groups like seniors, people with disabilities, and non-native speakers.

-Commuters and first-time visitors have different needs depending on trip purpose and familiarity.

-During transfers, riders face app overload and must quickly decide between various tools and transportation options.

Lead User Research

Brainstorming

Help team understand user

Pain Points

Lack Awareness of Existing Tools (Visitors)

Over 20 local transit tools offer real-time alerts and accessibility info, yet 87% of first-time visitors don’t know they exist.

Dynamic Needs (Local's Familiarity and Trip Type)

Users’ priorities change based on trip type and system familiarity—visitors need app guidance, while commuters want real-time updates.

High Support Demand

(Vulnerable Groups)

67% of inquiries come from riders with disabilities, seniors, or non-native speakers—highlighting the need for inclusive design.

Design Solution

1

Peronalization

1

Personalization

Customized Based on User Profile

2

Trip Specified Recommendation

Real time AI suggestions catering to different trip type

3

Accessible & Multilingual

Design Considerations

AI

AI

UX

UX

Inclusiveness

Inclusiveness

1

How AI Learns About the User?

Staged onboarding balances trust and personalization—starting light, growing smarter with use.

2

How to design for long-term engagement?

By identifying user types and their evolving needs, I created tailored journeys that adapt across familiarity levels and everyday transit scenarios.

Lead User Research

Persona

Brainstorming

2

How to design for long-term engagement?

By identifying user types and their evolving needs, I created tailored journeys that adapt across familiarity levels and everyday transit scenarios.

3

Right Prompts, Right Time

I designed contextual, scenario-based prompts using progressive input types—matching users’ mental models and helping the AI deliver focused, visual responses.


“Here are local apps!”

You Prioritize time!!!
So I list out time, cost and walking distance for you!


Are you going to school by bus as usual?
Hold On! A bus stop will change along your route!

Interactive Demo & Developer Handoff

Interactive Demo
& Developer Handoff

VUI

Prompt Structure Design

Technical MVP Development

Iteration based on usability testing

The design went through five iterations, guided by MVP check-ins with the sponsor, usability testing with 8 users, technical feasibility reviews, and feedback from domain experts.

Reflection:

This process taught me the importance of balancing user needs, technical constraints, and stakeholder goals. Iterating with real feedback helped me think critically, prioritize effectively, and stay adaptable in a fast-changing context.

Work under limited resource

With one engineer, I built a lightweight design system and clear docs to speed up handoff and development.

Reflection:

This experience taught me to think critically about scalability and communication—design isn’t just about visuals, but about enabling smooth collaboration under constraints.