With all the talk of AI-native telcos, it’s important to remember that to be AI-native, first you have to be cloud-native. What does that mean? It means there are very few operators that are cloud-native, so their perspective on being AI-native isn’t necessarily grounded in current reality. That said, there are a few cloud-native operators, so for perspective on AI-native, maybe best to talk to them. So we did. RCR Wireless News caught up with Rahul Atri; Atri was a managing director with Rakuten Mobile, left for Tech Mahindra, and rejoined Rakuten Symphony as a managing director and president of the OSS business.
The Rakuten Symphony stand in Barcelona, and our conversation with Atri, was sort of a microcosm for Rakuten. It’s all about fast. Think fast, talk fast, do fast. Which is pretty well the point of this idea of AI-native—get data fast, use AI to parse it fast, turn it into action fast.
Recalling his first run at Rakuten Mobile, Atri talked about the starting point where teams where grouped as radio or core or transport, whereas he looked after automation. Given that the big idea was to automate everything that could be automated, he said he had everyone first join the automation team, then re-split into more focused functional groups.
As to what AI-native means to him, “Cloud and data are the basis.” And that basis has to align with beneficial use cases, otherwise it’s just an exercise in moving data around various clouds. “Native to me is when you have a problem in your network or something day-to-day, the first thing you think is, can I solve it with AI or not?” Can a rules-based approach to AI that uses data in the cloud be used to solve a problem? “That’s what AI-native means to me.”
Atri was talking a bit philosophically about how product managers have to deconstruct problems into opportunities, then determine if those opportunities are pie in the sky or mapped directly to a problem that needs to be solved. He was getting into the notion that the arc of 5G started around bold new use cases that would lead to revenue growth and more effective network monetization, but contrasting that with the idea of doing things more practically and more efficiently.
Then he got a Viber call. He looked at his public relations person. “This is Mikitani-san. I have to take this.” And he did. And for a brief moment I saw the CEO of Rakuten Group, and two other people I couldn’t immediately identify, pop up on the screen of Atri’s phone. And he was off. Fast. Back to microcosms, that’s indicative of the culture that underlies Rakuten Mobile and Rakuten Symphony. I have a feeling there aren’t a lot of meetings about meetings or calendar invites flying through the organization.
Also in our conversation, Atri got further into his focus on identifying the problem and desired outcome then reversing through the solution, rather than the opposite approach of figuring out how to do something without necessarily knowing why you’re doing it. He talked about having to measure everything to understand whether there was a net benefit. Reminded me of Peter Drucker, one of your shining stars in the sky of management theory. Drucker said a lot of smart things over a long period of time. Here are a few that feel relevant:
- “Efficiency is doing the thing right. Effectiveness is doing the right thing.”
- “There is nothing quite so useless as doing with great efficiency something that should not be done at all.”
- And, “If you want something new, you have to stop doing something old.”
Rakuten Symphony had a good number of announcements released during MWC. The ones that hit me the most were the demo of an AI/ML model that resides in the RAN Intelligent Controller (RIC) that aggregates xApp and rApp intelligence from multiple cell sites then turns down or off particular elements when they’re not needed. The demo shows a potential 25% energy savings based on knowing what the network is doing so you can know what it doesn’t need to be doing—see Drucker bullet number two.
The other announcement was a collaboration between Rakuten Symphony and zTouch Networks combining the latter’s rApp/xApp portfolio with Symphony’s RIC and Service Management and Orchestration platforms. The point here is to use AI to enable the RIC and SMO to do “intent-based orchestration, [large language models] for zero-touch user interface, profit-aware traffic prioritization, and energy efficiency.” To summarize all of that, do things fast while saving money you otherwise would’ve spent and make money you otherwise wouldn’t have.
One more from Drucker while we’re all still here: “The best way to predict the future is to create it.”
