In this session, we will present practical case studies of implementing generative AI in Yahoo! JAPAN Flea Market. We will share insights gained through our step-by-step approach, starting from “product description generation,” progressing to “product title generation,” and ultimately developing the capability to “generate product information from a single image.” We will discuss our user value-centered approach rather than technology-driven implementation, establishing evaluation methods, and decision-making within practical constraints—all practical learnings from the field. For those working on generative AI projects, we will provide concrete tips on how to implement while maintaining a balance between technology and value.



"Yahoo! JAPAN Flea Market's" Use of Generative AI: Case Study of "Easy AI Listing" Generating Item Information from an Image
Speaker

Takanishi Yukiko / LY Corporation
Commerce Company Reuse Services Group Product Division 1 Reuse Labo Office Insight Team
Joined Yahoo Japan Corporation as a new graduate in 2012. I have experience as a Web, iOS, and API development engineer, and have worked as an engineering team leader. Currently, I lead the Insight Team at Yahoo! JAPAN Flea Market's Reuse Labo, exploring new business models for reuse services. I am also a member of the AI Promotion Team, working on implementing and utilizing generative AI technologies for reuse services.

Takeuchi Risa / LY Corporation
Commerce Company,Reuse Services Group, Product Division 1, Fleamarket Planning Department, Planning Team 1
I'm in charge of product planning for Yahoo! JAPAN Flea Market. My favorite hobby is collecting vintage items and works by artists.