Yen Dang - Product designer

Elevating Trust with AI-Powered Car Inspections

This project aimed to bring transparency to the used car market by incorporating advanced imaging and AI technology. Through a user-centered design approach, we created a streamlined experience on Twinner's platform, enabling sellers to digitalise and list their cars quickly and attract more buyers with professional and reliable listings.

My contribution as the first in-house product designer

Initiated Regular Testing Process

Collaborated with the process manager and content editor to establish a standardized approach and templates for various UX testing and research methods. This new framework empowers team members, including project managers, to independently conduct research as needed.

Enhance UX/UI of the platform

For each new feature on the roadmap, I developed the user flow, ideation, wireframes, high-fidelity designs, and conducted user testing. I primarily used platform like UsabilityHub and UserTesting for structured feedback and often employed hallway testing for quick, immediate insights.

Facilitate Product Discovery

On Mondays, our cross-functional team—including marketing, customer service, business, designer and engineering—meets for a brief sync to share feedback and ideas on the current product. I helped to facilitate the product discovery process. Ideas and insights are grouped by topic, prioritized, and added to the roadmap to guide our efforts in enhancing the service.

User journey mapping
Test setup with UsabilityHub
Interaction heatmap
Scan QR code to provide feedback
Notes on user interview

Let take a closer look at some specific tasks that I worked on

Enhancing and Expanding Features on the Twinner Platform

After having their car scanned at a Twinner Station, users can log into the Twinner Platform to access a detailed report on their car's condition, along with other valuable tools to support the car sale process.

Approach and process:

Discover: Dive deep into understanding user needs, goals, and pain points.
Ideate: Brainstorm creative solutions based on user insights and business requirements.
Design & Prototype: Translate ideas into detailed designs and interactive prototype
Test & Iterate: Gather user feedback through testing sessions, refine designs based on collected insights.
Deliver: Finalize designs & documentation for development. Work closely with developers & QA to ensure design intent is maintained.

Optimise the booking process to increase the conversion rate

Problem: Conversion rates were lower than expected, with significant drop-offs at each of the three steps in the booking process.

Goal: Streamline the booking experience to reduce friction and increase user completions.

Problem: Conversion rates were lower than expected, with significant drop-offs at each of the three steps in the booking process.

Goal: Streamline the booking experience to reduce friction and increase user completions.

Through testing, we identified several user preferences:

Location visibility is crucial for decision-making.
Week view was the preferred layout for date selection.
Payment options needed more visibility for added clarity and assurance

Through testing, we identified several user preferences:

Location visibility is crucial for decision-making.
Week view was the preferred layout for date selection.
Payment options needed more visibility for added clarity and assurance

Based on these findings, I redesigned the booking process to address user pain points:

Unified booking page: Consolidated the booking steps into a single page to minimize clicks and streamline flow.
Location info at the top: Positioned location of the station prominently at the top for easy access.
Week View: Integrated a week view calendar to simplify date and time selection.
Visible payment option: Added recognizable payment logos at checkout for increased trust.

Based on these findings, I redesigned the booking process to address user pain points:

Unified booking page: Consolidated the booking steps into a single page to minimize clicks and streamline flow.
Location info at the top: Positioned location of the station prominently at the top for easy access.
Week View: Integrated a week view calendar to simplify date and time selection.
Visible payment option: Added recognizable payment logos at checkout for increased trust.

Outcome

The new design improved usability, and users completed the booking process more quickly. However, we couldn’t measure a direct impact on conversion due to a significant marketing campaign that had previously run on the old version, which we couldn’t replicate for the updated version. This outcome suggests that external factors, such as campaign influence, pricing, and user motivation, may also play a critical role in conversion rates.

AI-powered car inspection process

Context: Twinner’s inspection process combines AI with human expertise to detect hard-to-spot damages. While AI identifies potential issues, inspectors review and confirm findings. An iPad app with augmented reality (AR) lets inspectors scan the car, instantly highlighting damage points.

Conetxt: Twinner’s inspection process combines AI with human expertise to detect hard-to-spot damages. While AI identifies potential issues, inspectors review and confirm findings. An iPad app with augmented reality (AR) lets inspectors scan the car, instantly highlighting damage points.

My Approach: I observed inspectors using the current app and conducted user interviews to gather feedback. I created an interactive prototype in Protopie, using the Wizard of Oz method to test AR interactions, and iterated based on inspector feedback.

My Approach: I observed inspectors using the current app and conducted user interviews to gather feedback. I created an interactive prototype in Protopie, using the Wizard of Oz method to test AR interactions, and iterated based on inspector feedback.

Outcome: The redesigned UI resolves previous usability issues, improving accessibility and streamlining the inspection process for greater efficiency.

Designed and crafted by me