Rakuten and OpenAI Partnership: Leveraging Generative AI for Customer Insights

Rakuten integrates OpenAI APIs to unlock value from ecosystem data

Rakuten is leveraging OpenAI's models to transform its vast repository of transactional, behavioral, and internal business data into actionable customer and business value. By combining these data assets with generative AI, Rakuten aims to transition into an "AI empowerment company," focusing on automating customer interactions and extracting structured insights from unstructured data.

Technical implementations and use cases

Rakuten utilizes a combination of OpenAI's APIs, Retrieval-Augmented Generation (RAG), and Code Interpreter to process complex data sources, including internal knowledge bases and unstructured documents like PDFs and Word files.

Automated Customer Service

Rakuten has reduced customer service response times from several days to automatic responses by implementing RAG on its internal knowledge base via OpenAI's API. This shift has significantly increased operational efficiency and user support speed.

Structured User Review Summarization

To improve the shopping experience, Rakuten is developing a feature that uses AI to extract key topics and summarize hundreds of user reviews. This allows customers to access product information in a structured format rather than manually reading through individual reviews.

B2B Knowledge Retrieval

Rakuten's B2B consultants use generative AI to derive actionable insights from company data, providing merchants and enterprises with market analyses and sales trends to help them compete in the digital landscape.

Security, Privacy, and Trust in the Japanese Market

Rakuten prioritizes security and privacy guardrails to maintain user and client trust, which is viewed as a critical component for innovation within the Japanese regulatory and social landscape. After a thorough due diligence process, Rakuten selected OpenAI as a partner based on the OpenAI's approach to security.

Future Directions: Real-time Voice and Vision

Rakuten plans to expand its AI capabilities by integrating real-time voice and vision features to process internal audio data for meeting action items, email communications, and multilingual translations.

Additionally, Rakuten is shifting toward conversational AI to better understand user preferences. According to Yusuke Kaji, General Manager of AI for Business at Rakuten Group, Inc., conversational inputs provide a more direct understanding of user pain points compared to traditional proxies like clicks and impressions:

"Clicks and impressions are just a proxy of a user’s preference. As we integrate LLMs more into the user experience, we can understand the user’s pain point or preference in a direct way through conversational inputs."

By capturing and managing this conversational data, Rakuten intends to further refine its product offerings and service quality.