Using car-sharing services has often required users to navigate maps and select rental locations to find their desired vehicle. However, Lotte Rental's subsidiary, Green Car, has simplified this process with its 'Lotte Rental G Car' service. By utilizing the 'G Car AI Reservation' feature, customers can simply state their desired vehicle or purpose, and the system will recommend and facilitate the reservation of suitable options.
In a recent interview, managers Seo Ji-yu and Choi Seung-pyo, who developed the service, emphasized that from the development stage, they focused on creating a 'more convenient reservation service' rather than just an AI service. 'AI is merely the technology underlying the convenience of the reservation process,' they stated.
The AX team, responsible for service development and validation, has consistently highlighted the importance of providing an experience that simplifies vehicle reservations rather than just showcasing AI technology. Seo noted, 'If customers feel they have 'experienced technology,' it means the planning has failed. The experience they should take away is simply that they rented a car more easily and quickly.'
They view the car-sharing model as akin to the process of explaining desired products to staff in a physical rental location and receiving recommendations. Just as customers can describe their needs to staff, the app allows users to articulate their requirements in natural language without having to manipulate search parameters on the screen.
Choi remarked, 'Customers often have clear requests, such as needing a hybrid car or wanting to try the latest BMW Mini, but translating these into screen commands has been the customer's burden. The starting point of our planning was to bring the experience of having the right vehicle selected based on who is using it and for what purpose into the app.'
One of the most challenging tasks has been reducing 'hallucination' in AI responses. Seo explained, 'We have established a system to verify AI-generated answers, tracking the ways and types of incorrect responses to improve answer quality. The AI verifies its own answers and analyzes the types and patterns of errors to enhance response quality.'
G Car plans to enhance its personalized services through AI development. In the long term, the goal is to evolve into a service that suggests necessary vehicles based on users' typical patterns without requiring them to explain their needs each time. Choi stated, 'Ultimately, our aim is to transition from a 'car rental service' to a service that designs mobility for users. The G Car AI reservation is the first step in that direction.'
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.