Edge cloud will enhance Japan’s already impressive track record of customer service and innovation, executives from Google Cloud and Rakuten Cloud have said. Having jointly supported a global retailer with more than 40,000 stores worldwide, the organizations are primed to replicate successful global outcomes from the “highly strategic” Rakuten Cloud-Google Distributed Cloud partnership within the Japanese enterprise.
All this and more was discussed on-stage at Rakuten AI Optimism – the Rakuten Group’s flagship conference hosted at the PACIFICO Yokohama Convention Center – where I had the pleasure of hosting Vijay Tewari, the Director of Product Management for Google Cloud Platform and Anirban Chakravartti (Oni), Senior Vice President of Enterprise Cloud Sales for the Cloud Business Unit (Rakuten Cloud) of Rakuten Symphony.
We discussed why Rakuten Cloud and Google Distributed Cloud have successfully supported customers across the world with their edge infrastructure modernization – and why there is massive potential for transformation in Japan.
Edge cloud has a broad range of use cases. Think any business or industry with a small footprint at an ‘edge’ location - such as an individual store - but with thousands of these sites deployed in locations across the world. En masse, these ‘edges’ comprise an enormous footprint, and individually need to be self-sustaining, reliable and able to provide the required infrastructure to deliver AI-driven applications.
The Google-Rakuten partnership in edge cloud has already delivered successful outcomes across numerous industries; Google Distributed Cloud – which runs on Rakuten Cloud’s Software-Defined Storage – is ensuring applications at the edge have the required speed, intelligence and processing power to deliver localized connectivity and storage, on-premise. This is allowing enterprises to deploy, operate and manage AI-based applications at the edge.
Vijay kicked off the discussion by putting into context the need for edge cloud – and why market opportunities are only set to increase because of an increased need for rapid data processing. “I'd start by first asking: why does the edge exist? Why do we need computing, storage and infrastructure at the edge? As the cloud evolved, it became abundantly clear that as applications run, they need network proximity to where the data is being generated - where the customer is interacting with the system. Whether it's a retail store, a manufacturing unit, a fast-food restaurant, you require infrastructure at the edge, close to where the action is happening.”
Oni echoed Vijay’s take. “We both see a lot of opportunities at the edge,” he said. “AI-driven applications are coming. What we are providing is whatever you are doing at the cloud, at the hyperscale level, you can get the same features at the edge level, which drives efficiency, productivity, and a better customer experience.”
The retail industry is ripe for transformation through edge cloud, Vijay said. There are often instances where workloads such as video streams would otherwise have to be sent to a public cloud, processed there and then sent back, requiring expensive network bandwidth. Instead of doing all this processing in the cloud, models can initially be trained in the cloud, then brought to the edge, where inferencing can be done over live video streams through Google Distributed Cloud and Rakuten Cloud’s Software-Defined Storage.
“Think about a retail store which is trying to use automated inventory management through video feeds, or a fast-food restaurant which is trying to improve their loyalty program through a customer ordering experience,” Vijay said. “What about gathering video feeds from a security camera that surrounds a retail store to do AI-based image processing on that video stream? That's the kind of architecture that we are seeing evolve in this distributed environment - where you can do large-scale training, RADs and customization in the cloud, but the actual inferencing happens at the edge where the data is being generated.”
The discussion moved on to ways that edge cloud can complement Japan’s widely renowned commitment to customer service and stellar quality in a variety of products and industries. At a Group level, Rakuten has made clear the opportunities that exist in Japan in terms of AI adoption – both Vijay and Oni agreed that edge cloud represents a significant next step in realising the opportunities around AI.
“Rakuten is deeply rooted in Japan - we understand the market very well,” Oni said. “We have learned a lot in the process of our global deployments and can replicate that success in Japan. Japan has been at the forefront of technology adoption and automation. Customer satisfaction is a very, very high priority as far as any organization here is concerned. Many Japanese organizations have a very large footprint outside of Japan, and that means there is a massive opportunity in Japan with the footprint of retail stores, fast moving consumer goods and manufacturing.”
Vijay echoed this sentiment. “Our infrastructure can enable Japanese enterprises to build richer applications so they can provide even better service for their customers. This infrastructure allows Japanese customers to bring innovation which they are known for, so that they can build these rich set of applications which solve unmet customer needs. Retail, manufacturing and fast-food restaurants are a great set of customer verticals that Google Distributed Cloud and Rakuten Cloud can actually bring to bear.”
Google’s ability to deliver edge cloud at scale comes from “years” of innovation, mainly though infrastructure and learnings from the public cloud which are now being deployed at the edge, largely due to Rakuten Cloud’s own innovation, Vijay said. “Google has vast experience in running infrastructure. We run infrastructure in our data centers at scale which are vastly beyond what most customers can dream of running - it is massive. We have learned over decades about what it takes to build and operate this infrastructure.
“We know how to statistically model how errors happen. We know how to build a system that can respond to errors and updates. What we are doing, along with Rakuten Cloud, is bringing a platform which bakes in all these learnings, and most importantly, we are taking the responsibility of operating this infrastructure. That takes away all the challenges that customers generally have in operating this kind of infrastructure at speed. You get Google's expertise, and you get the performance that comes by virtue of Rakuten Cloud Software-Defined Storage, which is baked into the solution.
“I'm very excited to see what Japanese customers will do once they have this infrastructure in their hands.”
