Join Kailem Anderson from Blue Planet as he discusses the future of network automation and its critical role in building autonomous networks. Learn about the evolution from partial to highly autonomous operations, how AI-driven solutions are reshaping the telecom landscape, and the benefits of breaking down organizational silos. With demonstrations on AI-based assurance, network as a service, and end-to-end network slicing, Kailem shares how Blue Planet is enabling the next generation of intelligent, self-healing networks.
Alejandro Piñero:
All right, welcome everyone once again to the FNTV studio right here in MWC, Barcelona. Well, we're entering day two, and once again, I have a great guest with me, Kailem Anderson from Blue Planet, who's going to talk to us about network automation. Kailem, welcome. Thanks for joining us today.
Kailem Anderson:
Thank you for having me.
Alejandro Piñero:
Now, Kailem, the objective of network automation has been there for a while. Of course, the technology is advancing, so can you share with us why that is so important and why that's an objective for so many CSPs?
Kailem Anderson:
Absolutely. If you look at autonomous networks, it's really about driving process and organizational automation across sort of the three organizations that sit within a telco, planning, fulfillment, and assurance, and really driving the inefficiencies that come out with those operational silos, passing information back and forth, and stitching together those businesses with process. Look at really what it brings autonomous operations. It's about driving self-healing, self-optimizing, self-organizing of the network across these functions, breaking down those silos.
What's the benefit of that? Very simply, well, the first one is there's a lot of passing information back and forth between those organizations, so it eliminates human error, right, fat-fingering configurations. Roughly 80% of network problems are related to humans. When you use a machine to automate human processes, you deal with the issues. The second one, very, very clearly, is customer experience, right? By being able to drive automation in closed-loop actions, you can deliver a better customer experience by being able to get to root cause and fix the issues in an automated way.
Some of the challenges with regards to implementing autonomous networks, very simply, it's not just about the technology, right? Technology is one piece of the puzzle, but if you don't go fix your organizational processes and the people, break down the organizational silos, then you don't truly get the benefit. So, addressing the technology with the process efficiencies that need to happen, and the people in organizational silos exist, is really the key elements in achieving the benefits of this technology.
Alejandro Piñero:
Got it. And as I understand it, the objective is to get to that level four, right, the highly autonomous network. What's the difference between level three and level four? What's that next step up?
Kailem Anderson:
That's a good question. Let me start with level two, because a lot of people are still in level two, which is partial automation, and that's really about automating a particular function. Take a NOC operator as an example. He gets an alarm, you want to do something with it. I might go to get more information, or I may reboot a server on that. It's usually using very basic process automation, if/then, and/or, yes/no rules, right?
If you go up to level three, conditional automation, really what that's about is starting to leverage data models in your automation, being able to bring in things like intent, declarative-based models where you have a state and a relationship between your fulfillment and assurance functions. Really starts focusing on the automation of services, whether that's an optical Ethernet/IP or business type service, and being able to understand where things need to change to be able to automate.
Level four, or highly, highly autonomous operations, is really that next step. It's the nirvana of our industry, and it's really about leveraging AI, AIOps, bringing in a predictive type capability, being able to program the network based on a prediction that you see in the network. Most importantly, it's multilayer. It's got intelligence, so say there's an issue at the optical layer, right? You have a fiber cut, then it knows what the impact is, say, at the higher layer system, such as IP in terms of congestion, and knows how to optimize across both.
Alejandro Piñero:
Excellent. And where do you guys fit into that puzzle then? I assume you're there to partner and to bring them to that nirvana, as you say?
Kailem Anderson:
Absolutely. Blue Planet, really over the last five to seven years, has really been bringing a suite of products together to go address this autonomous network issue, spanning planning, inventory, orchestration, and services assurance. And what we've done with those capabilities is build them on a unified cloud native platform and a unified data model that spans across all these areas. Why is that important? Because it maintains a state between planning, orchestration, and assurance, so as something changes, you actually know the impact of it through the complete lifecycle.
The final piece we've been doing is bringing in AI, AIOps, and our GenAI agentic capabilities through what we call an AI studio or a bring-your-own AI capability. So, we bring our own agents to understand that lifecycle from fulfillment to assurance, and when things change, how a network needs to be organized, and we allow our partners and customers to bring their own agents and apply it against our AI. So, a combination of having a suite of lifecycle products that are stitched together consistently, and then being able to apply AI in different combinations and permutations, really allows us to deliver on this vision.
Alejandro Piñero:
Brilliant. So, here we are in Barcelona. Before we wrap up here, I'd love to hear what you are showing at the show and what people can expect.
Kailem Anderson:
Yeah, well, thank you for asking. We have three key demonstrations we're showing. The first one is AI-based assurance, where we basically showcase this concept of bring-your-own AI or the democratization of AI, where you can bring your agents, apply it against our data set, and we'll show our agents are being able to deliver a broad range of use cases. The second use case, which I'm very passionate about, is network as a service, being able to show how a suite of automation products, delivered in a hierarchical model with APIs northbound and southbound, and being able to apply it against a unified data model can achieve a broad range of use cases to this concept of autonomous networks.
The final one, we're at a mobile event, so we wanted to show something mobile-related, so we are actually showcasing our end-to-end network slicing capabilities. Consider that almost an instantiation of the network-as-a-service type scenario where we show an end-to-end slice and then how we sub-slice spanning Transport, Packet Core, and RAN to deliver a unique experience that can be monetized back to the telco.
Alejandro Piñero:
Brilliant. Well, Kailem, thank you so much once again for coming over and talking to us about network automation. You've certainly educated me, and I'm sure a lot of folks out there, and best of luck this week in Barcelona.
Kailem Anderson:
Great. Thank you for having me.