AI Feature Design for ‘KakaoT'

Back

AI Feature Design for ‘KakaoT'

Designed AI-powered features to improve the taxi-hailing experience in Kakao T by reducing uncertainty throughout the user journey. Through review analysis, surveys, interviews, and usability testing, we identified key pain points around ride failures and waiting time, then proposed AI-driven guidance that helps users make more informed decisions before, during, and after requesting a taxi.

Year

2026

Type

Redesign Project

Role

PM, UX/UI Designer

Skills

Desk Research, User Research, AI Feature Design, Interaction Design, Prototyping, Usability Testing

Problem

Although Kakao T has significantly improved its dispatch speed and success rate, many users still experience frustration due to the lack of transparency throughout the taxi-hailing process. Users struggle to understand why ride requests fail, how long they will wait, or where they should stand to increase their chances of getting matched. As a result, they repeatedly retry requests, switch to other apps, or rely on guesswork when relocating.

To reduce uncertainty during the taxi-hailing experience, we designed three AI-powered features that provide actionable guidance throughout the journey:

AI Pickup Spot Recommendation suggests the most efficient pickup location based on road conditions, traffic direction, and legal stopping areas.
AI Dispatch Prediction estimates waiting time and dispatch success probability before users request a ride.
AI Re-request Recommendation analyzes failed requests and recommends better nearby pickup locations with walking directions to improve the likelihood of a successful match.

These features help users make informed decisions, reduce unnecessary retries, and create a more transparent and predictable taxi-hailing experience.

To reduce uncertainty during the taxi-hailing experience, we designed three AI-powered features that provide actionable guidance throughout the journey:

AI Pickup Spot Recommendation suggests the most efficient pickup location based on road conditions, traffic direction, and legal stopping areas.
AI Dispatch Prediction estimates waiting time and dispatch success probability before users request a ride.
AI Re-request Recommendation analyzes failed requests and recommends better nearby pickup locations with walking directions to improve the likelihood of a successful match.

These features help users make informed decisions, reduce unnecessary retries, and create a more transparent and predictable taxi-hailing experience.

Solution

Shot of a smartphone on a grey background