Yani Boyadzhieva, TelecomTV (00:04):
You are watching telecom tv. We are joining from Future Net World 2024 in London. Now I'm delighted to be joined by Laurent Laboucher, group CTO and SVP of Orange Innovation Networks. Hi Laurent. Thank you for joining us. Hello. So first off, how important is automation for network operators and where is Orange integrating automated processes?

Laurent Laboucher, Orange Innovation Networks (00:28):
Yes, so network automation for us is extremely important. We have started this journey already years ago. We are addressing different aspects of network automation, very focused on the way we do change. Change is about how you can bring DevOps and DevSecOps into the network operations and this is a big challenge as if we're progressing. In order to make it very concrete, we have implemented what we call a network integration factory, which is in fact instantiated in the different countries today and all the different network integration factories are working together in order to automate testing, integration and then deployment of network functions. There is a second part of network automation, which is also extremely interesting. It's about how you can leverage all the data that we get that we ingest from our network in order to automate some of the monitoring fault management anomaly detection processes. And we've started to do that with different countries. We already have some very interesting and very good first result. Now the question is how we can create this capacity at scale and how we can empower our network operating center to transform the operations into fully automated operations. It is a journey. So we are in the middle of the journey.

Yani Boyadzhieva, TelecomTV (02:40):
And you mentioned the six level model for autonomous networks based on TM forums scale and you mentioned that orange is on level two at the moment. What does this mean for the company?

Laurent Laboucher, Orange Innovation Networks (02:51):
Okay, so basically we are leveraging TM four and in T form there is this notion of measuring how autonomous the network are. We're using this scale of zero to five, which is very similar to autonomous autonomous car driving car. And we have assessed all the countries in all our geographies. So we have looked at all the different processes and looked at how automated we are today. So when I say we are at level two, it makes that so it's average, but it means that we have already started to implement the right processes. We have done the first key automation, for instance, the network integration factory is there, but we have not yet implemented AI at scale for the different use cases. In some areas we have implemented AI for root cause analysis for also predictive network managements but only in few places. So now we are trying to really bring those capabilities at scale, but then there are some challenges. The most important is to make all the data accessible. We call it data democracy and how we can bring the data in a data lake and hopefully we want to bring it to the cloud and then leverage that capability and implement with ML ops techniques and with our partners. We are working a lot with Google. We want to implement the different use cases to close the loops and bring automation.

Yani Boyadzhieva, TelecomTV (04:57):
You mentioned the importance of data. What are some other challenges to network automation?

Laurent Laboucher, Orange Innovation Networks (05:03):
One of the challenge is to bring the skills to upskill all our workforce at scale. So it's a lot of upskilling in my team within the different countries. It's also a lot of reskilling and we need to bring in the same teams, people who are very knowledgeable about data, data engineers, data scientists and people who are really extremely good at network operations, who know the business and who are ready to drive what is interesting in order, what can we automate. Something very interesting. During the past year we've watched very carefully the big progress made with generative AI and it's impressive to see the kind of capacity that we can get with generative ai and now we start to see how we can leverage in some very concrete use case to make our network very easy to interact with almost in using native natural language. So we are working on the first use cases to do that. So it's very exciting but at the same time very challenging to bring the right people together.

Yani Boyadzhieva, TelecomTV (06:31):
And during your session earlier on, you mentioned a backwards approach when it comes to use cases, finding the right use cases and monetizing the network. Can you share a bit more about this?

Laurent Laboucher, Orange Innovation Networks (06:42):
Thank you Yani. I like to see that as there are two sides of the coin. We are very focused on a automat the engine. But the reason is why are we doing that? We are doing that because we want to transform ourselves into a platform. We want to monetize our network. We want to find the right use cases where we want to be very, very focused. So as you say, we need to work backwards. That means that we don't want to bring, let's say all the bells and whistles at once and say, okay, now all the developers, you can use it. That doesn't work at all. What we want is really to be focused on key problems. One example is how we can leverage identity API in the network in order to help banks to mitigate fraud. Another example is how we can use the short latency of some networks on 5G, five GS, A in order to address some very specific industrial problems. Like for instance, if you need to control a GV in a factory, if you need to control the cranes in a harbor, how can we be very focused on those very specific use cases? We don't want to boil the ocean, we want to select a few of them and be very focused.

Yani Boyadzhieva, TelecomTV (08:25):
How important is it for you to partner with the right existing players?

Laurent Laboucher, Orange Innovation Networks (08:29):
We will not make it alone. It's a industry journey. So first of all, we need to leverage our peers. We need to work together. At the same time we are competitors, but we are co competitors and we share the same foundation on the foundation on the basics. We need absolutely to create scale. We need to create scale on the telco cloud. We need to create scale on the basic network APIs and we need also to bring other partners in the ecosystem. I strongly believe that hyperscalers partners are very important. We need to leverage them at different level on the telco cloud for some workloads. We need also to work with them on ai, how we can use all their very powerful capacities. We are working very closely with Google as an example, but we are also working with AWS in order to deploy 5G cloud network. So we want to work with all of them and each cloud provider today has some advantage and we want to take the best of each one. I think with hyperscalers we also have to play in their marketplace because they have a very strong and very powerful marketplace where they can attract developers on some verticals. This is not our DNA and we can leverage their capacity and really work very smoothly with their marketplace.

Yani Boyadzhieva, TelecomTV (10:14):
That makes perfect sense. Laurent, thank you very much for speaking with us. Thank you very much.

Laurent Laboucher, Orange Innovation Networks (10:18):
Thank you.