As Rakuten expanded into an impressive ecosystem of more than 70 businesses across e-commerce, travel, banking and more, maintaining speed and productivity became a challenge. To solve this challenge, Izumi has led the production and development of a sophisticated AI platform. Under Izumi’s leadership, 84 chatbots (and counting) have been launched, and the company estimates around 83% of its human workload in e-commerce is now handled by AI. There are several bots in production for each business, to meet a variety of needs and demands.

Please tell us about your project and how you’re using IBM Watson in your business.
Rakuten now serves 100 million users in Japan and 1.4 billion users in more than 30 countries and regions, so we needed to leverage the advantages of AI. We have established four AI platforms using IBM technology including IBM Watson Assistant and Watson Discovery. Now customers can consult with virtual agents 24/7, with AI handling the equivalent of 15,000 hours of human work per month.

What makes an AI project successful?
It is essential that we put the customer first. If the solution does not solve the customer’s problem, they won’t be using the solution. On the basis of that “customer first mindset,” we always try to communicate with users to get feedback on how we can do better and what to improve about our solutions. In addition, motivating developers is crucial. New technology emerges and advances every day at unprecedented speed. It is critical that we create opportunities for developers to try and keep up with new technology.

What advice would you share with others who are considering using AI?
Business users sometimes think AI can do everything automatically if we input lots of data. But that’s not correct. We need to clarify what AI can do and where we can use AI to great effect.

She puts Watson Assistant to work

Dive deeper into your journey to AI and learn more about Watson Assistant