Rakuten

The Story Behind Rakuten

Rakuten Inc. is a global internet innovation group with over 70 businesses and more than 1 billion members worldwide. Its ecosystem spans e-commerce, banking, telecommunications, digital content, advertising, and more. Rakuten was founded in Tokyo in 1997. Its B2B2C e-commerce platform Rakuten Ichiba is the largest marketplace in Japan and among the world’s largest by sales. This company provides several services globally, including Rakuten Viber, Rakuten Advertising, Rakuten Intelligence, Rakuten VIKI, and many others.

When the client approached Zazmic, they needed technical help with one of Rakuten Advertising’s services products. Rakuten’s global advertising business includes its top-rated cashback program Rakuten Rewards, formerly known as EBATES. Rakuten Advertising has multiple AdTech Platforms, most notably its Affiliate, DSP, SSP, and Data businesses.

Business Challenges

The client faced a business challenge that required an experienced technical team to cope with it. Rakuten acquired several companies in the Adtech space using on-premise and legacy technology. The original systems and architecture were hard to monitor and needed to reach very high uptime KPIs. At the same time, a large amount of data that was logged daily needed to be processed very quickly. Thus, the main challenges for our team were the following:

• Real-time optimization performed by the platform needed quick response time and high uptime to achieve the performance required to meet the KPI needed by the business;
• Creating a new, more robust data pipeline to allow for better Analytics and Machine Learning;
• Modernizing the Real-Time Bidding (RTB) framework for one of Rakuten’s large Scale RTB Platforms for Display and Mobile.

Solutions Delivered

To address the challenges presented by the Rakuten company, our technical specialists developed a detailed plan and implemented the following steps:

• Rakuten replaced the legacy data pipeline and migrated to Compute Engine and Cloud Data Flow. This was done in record time and provided immediate cost saving, greater performance, and higher stability.
• Zazmic, in collaboration with the Google Cloud engineering team, helped Rakuten’s team migrate to the Google Cloud Platform and engineered a solution to modernize the legacy infrastructure.
• Migrating the legacy platform to Compute Engine, C and Data flow.
• Zazmic worked closely with Rakuten’s IT Technical operations to enable auto-scaling via Kubernetes and uptime by integrating Data flow.

Outcomes

Zazmic’s solution allowed Rakuten to complete modernization, achieving 99.9% uptime since migration by ensuring that the real-time ML Product has all the data necessary to perform. Further, the KPIs mandated by the business were met and improved. The solution also provided cost-effective storage for petabytes of data. Additionally, it enables better data science on a larger dataset, reducing the time to produce an effective model.