Ray Le Maistre, TelecomTV (00:07):
Andrew second time up here, but we do have some new people in the room. So if you could just tell us quickly who you are and where you're from.

Andy Burrell, Nokia (00:14):
Sure. And for those that weren't listening the first time, I'm still Andy Burrell. I'm still heading marketing for one of Nokia's businesses called Cloud and Network Services. It's a software based business, OSS, security and also core networks. And yeah, really glad to be here.

Ray Le Maistre, TelecomTV (00:29):
Okay. And before we're going to look at some of the, go into a little bit more detail on some of the things you talked about this morning, but just in general, what's Nokia's approach to helping telcos with their AI requirements? What are the main engagements that you are having with operators right now in terms of ai?

Andy Burrell, Nokia (00:51):
So this morning we spoke about sense, think, act, these three layers. That was our kind of framework. A lot of the focus there was on data governance, data quality and so on. I dunno if there is really a simple answer to that question because the nature of our engagements varies. I mean the industry is not homogenous. The discussions we have with a tier one CSP in the us, different to those that we might have with somebody in Middle East, Africa or Asia and so on. And there seems to be a sort of changing pattern. So a couple of years ago, energy was very, very hot. I mean, obviously there was crisis going on and spikes in prices. So that was the single topic that we were talking to most of our customers about. Before that we had security prompted by things that were going on in the world. It seems to sort of revolve customer experience at one point was the lead topic. A lot of it seems to come back to cost. It's slightly depressing, but anytime you can talk to a customer about cost reduction, suddenly the doors are open and the discussion gets a bit more serious. So it's really varied, I would say.

Ray Le Maistre, TelecomTV (02:04):
Okay. Now, one of the things you talked about this morning was this sort of the network nirvana related to automation. And we know that there are some operators that have put themselves quite far down the path. I imagine some of that is in particular parts of their operations where they've reached level four of level five, but you also talked about a figure, and I always like to hear money numbers talked about with an potential 800 million saving from full automation for telcos. Can we just break that number down and find out what that 800 million means? I would imagine that's probably the largest operators in the world going almost fully autonomous with their day-to-day operations.

Andy Burrell, Nokia (03:02):
Yeah, so I mean before we get into that, we've got a lady from SDL in the room actually. So this was some research we did together with SDL partners and also Charlotte Patrick Research. We had some discussions about the modeling and our perspective on it, some interviews with CSPs and so on. That $800 million headline figure is based on, in quotes, a typical service provider. What does that mean? It's one with revenues of 15 billion around about 30 odd million mobile subscribers, and I think it's like 12 million fixed subscribers. So you can argue is that representative or not, but it's not necessarily the biggest. It's not only China Mobile or Verizon that you could get those kind of savings for. And of course I think as a percentage you've got the same chance to make those savings. And then I guess the other thing, I should correct, the way that I worded it to keep it simple was 800 million savings. It's about 650 million I think in savings and 140 or whatever it is in incremental revenues as well

Ray Le Maistre, TelecomTV (04:15):
Of value and over what kind of time period here as well. Because

Andy Burrell, Nokia (04:20):
I mean the savings are on a per annum basis, and this is on the assumption that you move to level four, level five autonomy as TM forum define it in their maturity model.

Ray Le Maistre, TelecomTV (04:34):
Okay. Are these the kind of numbers that generate interaction that get the service providers of the world interested to find out more?

Andy Burrell, Nokia (04:46):
To a certain extent. I mean, it's a big enough number I think to attract some interest. People want to know where does that number come from? Just the discussion that we're having here. And yeah, what's realistic for me and my organization. And of course you can break that down into different domains where you think you can get quick wins, fault management as an example, or field engineering or something like that. But yeah, it's a big enough number to attract attention for sure.

Ray Le Maistre, TelecomTV (05:13):
Okay. And then again, earlier on you were talking about use cases, but you also talked about some specific customer engagements where some of your knockers technologies have been used in specific examples. Can you drill down into some of those a little bit more? Telephonica, Germany and Claro, I think were the two that you were able to name that were on your slide earlier.

Andy Burrell, Nokia (05:44):
Yeah. Okay. So those are both public. I mean, there's been press releases and video testimonials and so on about those. And the first one, Telefonica Germany, it was Malik Rao, the CTO there, who was sort of vocal about this. He's a bit of an advocate for SaaS. So this is our AVA energy efficiency solution deployed in a SaaS model, classic kind of approach that we have with pretty much all service providers. It starts with a discussion around what's the use case, what's the value for this? I mentioned about a couple of years ago how hot energy was as a topic and how everybody was feeling the pain classic sort of approach where we have to go and prove the technology works, will it actually make the savings that you are predicting the headline figure? There was like a 10% reduction in energy consumption and then going into scaling that across the whole of the network. And it's a really obvious thing, but when you're talking to the CTO organization, there's always concerns about quality and potential impacts and a natural reluctance I suppose, to take hands off and to let the system make changes to the network. But yeah, that's a journey that we've been on. And yeah, we're starting to see the results already.

Ray Le Maistre, TelecomTV (07:03):
And with Claro, that was the focus there was on security, which we haven't heard talked about a lot today, but it's one of the areas where machine learning in particular has been deployed for many years already.

Andy Burrell, Nokia (07:17):
Yeah, I mean it's really obvious in a way that security's important and has a lot of scope for getting benefits from ai. For some reason we kind of treat it specially and have separate forums where we discuss it with Clara Columbia. Again, that's a public case. I think the press release was out sometime earlier this year. It's applied to their 5G network, about 35 million subscribers on that network. And I think there's actually three sort of separate security solutions that we've deployed there. The first one is endpoint detection and response. So that's an agent that's deployed on the network functions to detect, as an example, malware or intrusion. Then there's a privileged privileged access management system. So that's monitoring what Clara's own employees are doing because insider threats, whether that's malicious, somebody trying to blackmail you or whatever, that's a big source of problems across all organizations. And then the last point, we had something called Cyber do, which is our extended detection response platform, and that sits above the endpoint detection and the privileged access management solution.

Ray Le Maistre, TelecomTV (08:37):
And finally, before we move on to the next session, can I ask what is next from Nokia in terms of its AI product development?

Andy Burrell, Nokia (08:46):
Wow. I think there's been a lot of hype about generative ai. There's been a few sort of atomic use cases where we've seen it applied to things like customer care and so on. A lot of use of copilot type functionality. I think we're still fairly early in that journey. I think one thing that I think is quite exciting as we spoke about data products. So the idea that you take the raw data, you curate the data, you make it usable in a form for a data scientist, I think there's a lot of scope for generative AI to empower data citizens to do some of the stuff that data scientists are doing today. And that's an obvious kind of efficiency gain, but also just increasing the possibilities for telcos. And we've always had the data, getting access to it, making it available. That's been the challenge and I think we're maybe on the point of making a step change in that, I would say.

Ray Le Maistre, TelecomTV (09:54):
Okay. Excellent. Well, Andy, thanks very much for helping us to open up this afternoon's proceedings. We're going to move on now to the first session of the afternoon, but give a round of applause, Andrew. Thanks.