OpenAI

Rakuten Leverages OpenAI API to Enhance Customer and Merchant Services


Executive Summary

Global technology company Rakuten is partnering with OpenAI to apply advanced AI models to its extensive data from over 1.8 billion members and 70+ services. Rakuten is using OpenAI's API, integrated with techniques like Retrieval-Augmented Generation (RAG), to automate customer service, develop new e-commerce features like review summarization, and provide data-driven insights to its B2B clients. The stated goal is to become an "AI empowerment company," amplifying the value of its data assets while maintaining a primary focus on security and privacy.

Key Takeaways

* Automated Customer Service: By using OpenAI's API with RAG on its internal knowledge base, Rakuten can now automatically and instantly respond to customer service tickets, drastically reducing wait times.

* Enhanced E-commerce Experience: The company is developing a feature to summarize large volumes of user reviews, allowing shoppers to understand product feedback in a more structured and efficient way.

* B2B Data Insights: Rakuten consultants use AI tools to analyze complex, unstructured data, providing merchants and enterprise clients with actionable insights like market analyses and sales trends.

* Security & Privacy Focus: Rakuten explicitly chose OpenAI after a thorough due diligence process, emphasizing that trust, security, and privacy are critical components of its AI strategy.

* Future Vision: The company plans to explore real-time voice and vision capabilities for internal productivity and integrate more conversational AI experiences to better understand user needs directly.

Strategic Importance

This initiative allows Rakuten to unlock significant value from its proprietary data ecosystem, improving operational efficiency and creating more personalized, data-driven experiences for both consumers and business clients. It positions Rakuten as a key enterprise adopter of generative AI, setting a precedent for securely deploying these technologies at scale.

Original article