How AI Reshaped Network Operations for a Telecom Provider

Marc LeBlanc (00:01.934)
This is Solving for Change, the podcast where you'll hear stories from business leaders and technology industry experts about how they executed bold business transformation in response to shifts in the market or advances in technology. In every episode, we'll explore real world strategies and technologies that fuel successful evolution. I'm your host this month, Marc LeBlanc.

Marc LeBlanc (00:34.734)
In this episode, we'll take a closer look at how modernizing infrastructure for Canadian telecommunication company had a direct impact on the organization's ability to stay competitive and an industry that thrives in innovation. We're very fortunate to have Hart Ripley here to talk to us today about what this looked like. Hart is the National Automation Lead and Solutions Architect for the Office of the CTO here at Maldia. With over 20 years of experience in IT, Hart has a wide breadth of experience while he's worked with

In many areas of IT, his passion and expertise is in building innovative custom solutions to tackle real business challenges. Always up to speed with the latest trends in the industry, he quickly spots the ones that have the potential to be the most impactful. Thanks for joining us today on Solvable for Change Heart.

Hart Ripley (01:24.732)
Thanks, Mark. Happy to be here.

Marc LeBlanc (01:30.232)
So tell me a little bit about the problem that this particular telecommunication company was having. What was it that got us involved to begin with?

Hart Ripley (01:41.203)
Mostly centered around the network operations, right? A lot of network operations tools for subscribers and kind of outages and there's a lot of hardware in the field, right? To really keep tabs on, make sure everything's working and...

They were long time users of NetCool Network Management, Operations Insight, those types of things like that, as well as SEV1. Some of these tools are deprecated. Some of these tools were being rolled into more of IBM's AIOps tooling, Watson AIOps. And really it was about consolidating that, but also giving them the ability to have more actionable insights, some better observability, some better visibility, but really understanding what's happening in the network, how to respond.

to it in more of a real -time fashion and less reactive.

Marc LeBlanc (02:33.506)
And what was fueling that? So I kind of take away that there is a bit of a disparate tooling ecosystem is shifting away from maybe what they needed. But where were they trying to get? What was the goal they needed to realize?

Hart Ripley (02:51.699)
They're trying to get to more of a consolidated way to action in the outages, but also look at trends in the network, right? So really seeing what, you know, busy end points, you know, when things are getting overloaded, where they're coming from, but being able to correlate that over time, right? A lot of the, you know, older network management solutions and insights were more focused on historical, right? And relied heavily on kind of human intervention.

Right? So let's look at a problem. Let's, get some folks involved. Let's, let's do some troubleshooting, some triage and then, you know, time to resolution is much longer. So with some of these AI ML models and approaches, right? They can turn that kind of meantime to resolution way down, right? They can get an idea where the source of the issue is coming from. They can drill down faster. You know, the technology and the solutions gives them the ability to do more automated response.

right, kick off workflows, know, run some automation jobs, those types of things like that if they want to. But the main thing is really just rolling that up and having the real insights in real time.

Marc LeBlanc (04:03.98)
Right, and what was the approach that was taken to get them closer to that goal? What were some of the processes or some of the tools that maybe were introduced as part of the solutioning?

Hart Ripley (04:15.217)
Yeah, so the big one was the adoption of the IBM Watson AIOps Cloud Pak.

Right? This is a containerized tool that really will take inputs from, know, SNMP inputs from other, you know, IBM tooling, including SEV1, which was the real main source of ingestion here. You have things like SNMP traps. You have other data sources that can send data to AIOps. Right? And the big driver for them was this is a containerized application. Right? It needs a container platform. Right? It's very much a cloud native, you know,

tool and solution and they already had the tools to feed in the network information. This was the piece of the puzzle they were missing was the overall observability, the insights piece, the AI operations platform to really give them a centralized view, but also give them the ability to do something with that information.

Instead of having a bunch of disparate tools, let's take all these tools. Now they can send all their data into one area, aggregate that data, and start seeing what are the trends we're seeing? What are the areas where we see some issues? When are issues starting to materialize? And get more of a forward -looking proactive approach into this tooling.

Overall, it was a containerized platform. It's a cloud pack and it really comes out of the box with all these kind of offerings. But with this telecom, what wasn't really familiar with was more of the containerized journey, right? They were focused a lot on virtualized workloads and they didn't have a container platform of choice. And that's where Mobia came in and to kind of help them with that journey.

Marc LeBlanc (06:08.574)
It's interesting, I think that's something we see quite a bit in our industry is that jump from virtualization up to a containerized platform. It's quite a jump, it's a different paradigm way of thinking. What was the approach to gain comfort for both the leadership and with the people that are gonna operate that? What were some of the steps that were taken to get everyone to good or to be okay and comfortable?

Hart Ripley (06:34.279)
Yeah, I think that's one of the areas where Mobia really adds value, right? We're not just looking to come in and kind of pitch something. We're looking to come in and work with the customer and help them understand. Yes, it's a journey. Yes, it's a new piece of technology, but you you have an operations team already. You have an infrastructure team, you know, largely you're already doing a lot of these things, right? It's not, it may seem like a black box initially, but it's really for us to kind of help and say,

this is a journey. There's some new tooling, there's some new graphical interfaces, there's some new command line tools potentially, there's probably a new user experience is the best way to put it. But at the end of the day, we will help you with those processes, we will help you understand what's required to manage it, what the life cycle looks like, and really try to make that as comfortable of a transition as possible.

And containerized journey in this case was for a specific use case.

You know, we're looking at the bigger picture. It's not about how do we run this cloud pack today? It's not, how do we get this working for the single use case? It's how can Mobi enable you to, you know, be able to scale down the road or be able to, you know, expand out your containerization requirements down the road. Right. So it's, it's very much a collaborative engagement. That's really how we work with these clients is yes, we want to help you meet your goals right now. We want to have you help you have some

success on your immediate requirements, but also not losing sight of, you're going to need to expand more on the container side down the road. Other off -the -shelf tooling, know, other open source tooling, other in -house development is going to drive the requirements for microservices, you know, other containerized outcomes, those types of things like that. So it's important not just to say, okay, we're going to build a container platform and we're going to run this today. It's how do we integrate this into our

Hart Ripley (08:37.983)
know, operational life cycle, our operations team, our infrastructure team, our, you know, response management, our SRE practices, all these types of things to make sure that it isn't a black box really to the organization, you know, and obviously the team that needs to support it and run it also is as enabled as possible, least day one, and then understanding what they can do with the platform and their investment.

Marc LeBlanc (09:05.205)
And as far as that operations team, there's a journey for them as well. You touched on a number of the roles or functions that might be required to support that team, or sorry, that platform. What's the path to get there? Are these typically new hires or is it upskilling? Is it a blend? Is it looking for partners that can maybe take on the operations? What was the approach that this particular customer went through?

Hart Ripley (09:31.557)
Yeah, we looked at it pretty holistically. was the customer didn't have any experience with containerization or running a containerized platform. So the last thing we wanted to do was drop in something that, you know, we documented, we did for them, and we kind of walked away at the end of the day, right? We had multiple parties here. You know, we talked about, you know, a vendor being IBM, right? Red Hat, OpenShift being the container platform of choice, very much aligned with IBM and Red Hat strategy.

and obviously the fact that they're under one umbrella that...

OpenShift is going to be the platform for any cloud packs that come from IBM. Everything's certified in OpenShift. Everything is versioned and you have a very clear picture of what the life cycle is for that cloud pack. Not to mention entitlements, subscriptions, those types of things are part of the cloud pack. And really one of the driving forces to adopt OpenShift to run these cloud packs, there's multiple advantages. You've got both vendors supporting it cohesively.

And that's one thing. So I think the big thing is you have internal folks that, know, a lot of the times there are systems administrators, Linux admins, know, infrastructure folks. They've done a lot of virtualization management, hypervisor management, those types of things. So it really comes to that individual approach where, you know, there's some upscaling required and...

You know, it's also important to make it a very personal thing. It's, you know, it's a learning experience for some of these folks. It's an opportunity to, you know, on the job, learn new skills and learn new things and, you know, add things to your kind of overall experience that, you know, can benefit the organization, can benefit you down the road as an individual. And you're really looking to find these stakeholders and find these champions because, you know, we, don't want to help enable technology.

Hart Ripley (11:32.698)
in these organizations if the individuals really aren't aligned with the goals. So the main thing for us is to make sure that they understand and they're on board with, we're here to help you, we're here to work together. We're not here to do it for you. We will enable you to do as much as you want hands on yourself. We will do collaborative engagements, collaborative delivery. Part of it, to your point, is there's operational management.

kind of day two, day three. And the big thing there is, you know, if you don't have a team that's, you know, run container platforms today or manage the lifecycle of a container platform, stuff like that, you know, you have your managed service providers, in which case, Mobiya, you know, we are doing managed Kubernetes today, managed OpenShift, right? So there's that option. It was really from day one trying to understand, you know, was this client...

looking towards outsourcing some of this type of stuff. We were looking to provide the help, the enablement, the operational continuity where we could. And in this case, we were very much aligned with, okay, you have some individuals that are gonna get upscaled in some areas.

but overall operational management as you scale out the number of clusters, other app use cases, those types of things like that, it's just, it's not gonna be a core competency or core function day one. So, let Mobia take the managed services, the ongoing versioning, the break fix, the proactive alerting and response, those types of things. And let's focus on allowing you to consume the cloud packs, the AIOps, intelligence, the

automated remediation, the other Intel that you've got.

Hart Ripley (13:24.473)
and is available to you by kind of going down this road. And, you know, we'll learn together. We'll go together down this journey and, you know, we'll continue to enable you. And, you know, we'll understand that, you know, at some point your use cases are going to expand, your footprints could expand, your team's going to expand. And, you know, we'll just, we'll just stay very aligned with you and understand how can we help, you know, what do you need? Let's, let's continue to do roadmap stuff for the future to understand where can we

help, where can we plug in, and as much as we want to help with enablement, really understanding what that team's going to look like down the road and kind of right now.

Marc LeBlanc (14:08.032)
Right, and you touched on something there. We have this platform in for very specific use case, but we know that when we bring in a platform like Kubernetes, it does expand. We do bring in other applications and workloads. And yes, you need help upfront maybe from a partner for the operations, but by coming up with a holistic approach, you're future -proofing your team and your platform to grow with you.

But I want to switch gears a little bit. talked about the platform, but really we want to dig into what role that AI tooling is playing in this. So this platform is to get ready for an AI tool. Maybe talk a little bit more specifically about what that tool is providing and how it's going to enable NOX to better prioritize those alerts and get it better in sight.

Hart Ripley (14:44.007)
Mm

Hart Ripley (14:54.267)
Yeah, for sure. So I touched on this a little bit earlier, but there's a lot of disparate tools initially where they had to go look at SNMP collectors for certain issues. They had to look at network, net cool operations management for certain things, SEV1 for certain other things. The point of AIOps is to collapse some of these older tools into more of a modern framework, right? Where we have...

we're taking in all this data, but what are we doing with the data? How are we correlating it? How are we aggregating it? How are we analyzing it? Right. And the knot that we have today is very much, you know, not used to kind of that automated remediation, not really used to.

being able to see too many trends ahead of time and be able to do that action bloat comes on them. So the point of the AI piece and the ML piece that's built into the AI ops is, you know, now we have all the data in one place. That's kind of point number one, right? Before we had data all over the place, we needed to get data into kind of one common platform from all our kind of sources and senders. And now we have the ability to say, okay, now we have built -in monitoring, right? Now we have, you know, a bunch of correlation.

have visibility, now we have trending, now we have kind of suggestive analysis to understand, you know, we're seeing these issues or these types of trends on these types of devices, right? And we're able to kind of build a better roadmap and a better understanding of what does our entire network footprint look like, you know, where are we seeing issues or congestion, where are we seeing, you know, even physical issues, port issues, subscriber issues, handoff issues, you know, as well as user experiences.

issues all kind of rolled up into that that one spot but it's not about

Hart Ripley (16:41.789)
you know, before having to go look at these tools and kind of trying to normalize what they're telling us. Now it's very much rolled up into the AI models to kind of say, well, here's what we're seeing, right? It's very visual. The dashboards are very customizable and they're producing, you know, obviously customized things if that's what the organization decides to do. But there's also a lot of built -in kind of actionable outputs, workflows, dashboards, visualizations to say,

Okay, I want to show my, you know, my executives kind of what our, you know, network health looks like or what our trends are over a certain period of time. Right. So not only can this benefit the NOC and the SRE teams, but we can say some of our executives who've, you know, supported these investments helped us kind of get some of this newer tooling in the house, in the door. You know, what are we able to see now?

You know, how far ahead can we look? You know, what's my network look like at any given time? How healthy is it? You know, where am I starting to see some issues, you know?

when we see a lot of very heavy workloads on the internet with some of these events and these sporting events and stuff like that, how's my network responding? Those types of things. So it allows us to not only roll it up to the operations team, but the tools itself will allow the different consumers a different type of user experience.

and really be able to say, we're not just going to look at this, we want to put it out there for different team members to kind of see in real time what's happening.

Marc LeBlanc (18:21.96)
You touched on a number of things there, number of small wins and bigger wins. I'm wondering if maybe you could highlight beyond some of the challenges we talked about having to bring in new technology, getting your operations team aligned. Were there any other challenges that we had to overcome as we worked through this solution with this company?

Hart Ripley (18:41.029)
Yeah, I think the one of the big things is, you we're talking about integrating a container platform, you know, and running, you know, a cloud pack on top of it, right? So there's a lot to unpack there, but, you know, container platforms have a lot of requirements, right? We're talking, well, there's still infrastructure, there's networking, there's storage, there's security, you know, there's traffic management.

There's user access control, user experience, onboarding, right? So there's a lot of stuff to consider, right? And that's where Mobia really adds the value to say, okay, you're going down this road for the first time, right? We're building out a production containerized platform, right? It is gonna run, you know, kind of an out of the box, off the shelf operator that's natively integrated with OpenShift, but we still need to make sure the platform is gonna be healthy, make sure the platform is going to have stability.

the storage is always available. I think that's one of the things that really stands out to me is that the more stuff you're running on a platform, especially kind of these multifaceted tools, there's a lot more requirements that need to be looked at day one. And we spend a lot of time in the architecture design phase because...

That's really key. We're not going to be able to successfully deploy, operate, manage a platform, you if we miss something in terms of the architecture and design or those specific integrations that are going to support, you know, the platform's ability to run, to run well, to scale, those types of things like that. So I think it's...

It's important that we focus on that holistically, but also understanding that we do have some operators on top. We do have other requirements within the infrastructure that need to support it.

Hart Ripley (20:29.253)
really it's about kind of looking at all that stuff in context and making it easy to kind of disseminate for all the individuals, but also for the organization to say, you know, okay, this is new, but you know, we're familiar with storage, we're familiar with networking, we're familiar with infrastructure, it's just, you know, now we have some different architecture, different design, and possibly a different consumption model. And really that's kind of the way we tee it up. We really try to make it less of

of an unknown or a big learning curve. There's definitely some pieces that will require a little bit more than others in terms of learning and adoption, but.

At the end of the day, a lot of the skill sets are transferable. And I think that's one of the big things that may get lost or not talked about enough when we're modernizing platforms is, you're a systems administrator. You understand networking, you understand permissions, you understand operating systems, you understand infrastructure to a point. Well, all those skills are very transferable in the modern platform containerized world. So just really trying to wrap that together and

it a little bit easier to to kind of move from from one to the other to understand kind of how some of that stuff stacks up.

Marc LeBlanc (21:48.726)
Right. I'm curious, as we rolled this in, you were working with the teams planning, doing their design, doing their architecture, doing the upskilling. Has the organization seen any impact on the culture within those teams as part of this journey?

Hart Ripley (22:06.407)
A bit, yeah.

For sure. The way that we approach a lot of these is that we want to follow as much of a DevOps methodology approach as we can. We're talking about modern platforms, we're talking about containerization, microservices, code, images, those types of things. So we are going to deliver in a way that not only are we going to enable the team on the platform and the operations side, but we're going to give them the keys and

the training and the enablement to really understand how do we keep this working? How do we do this all as code? How do we take, we're not copying scripts around, we're not copying files around every time we deploy a cluster, we're going to do it in a very declarative fashion. Just by the nature of how we talk about this, how we deliver it, how we go through workshops with these teams.

It really starts to settle in to say, okay, we're using source code management. Everything we do is a repository. We have a set of code. You know, here's our starting point. Here's our blueprint, if you want to call it that. And we really obviously take a step back and look at the, you know, the Git.

pieces and really how to use Git, how to use branching, right? How do different repositories work? How does permissions work? You know, how do you do promotion of code, those types of things, just the fundamental pieces to say, you know, we're not going to be doing this in, you know, SCP and copying documentation around. We're not going to be documenting this in Word. We're going to be documenting this as code and just, you know, everything goes in a repository. We make a change to a document or an artifact. It happens in the code repository, right?

Hart Ripley (23:53.149)
and the team that we're working with is alongside our delivery team is going to have access to that code repository. We're to do things consistently. It's going to take a little bit of time to and a little bit of change up front. But once you get the team on board, people start to see the light and really understand that, I can look back and see when this used to work. What was the code base then? Can I revert to that code base? What were the changes made? Who made the change? When did they make the change?

those types of things. So there's a lot of advantages to kind of going down that approach and really adopting that, you know, kind of that agile methodology and really the those DevOps focuses. And that's what we're really trying to teach is just more of that overall workflow, that culture. And, you know, I think by kind of leading with that to your point is, you know, we're trying to build a culture, you know, that, you know, we have in Mobea, but we're trying to also enable that in our customer environments.

and especially individuals and teams.

Marc LeBlanc (24:56.874)
Very true. Why don't I shift gears a little bit and start talking about kind of the value back to the business and what that future roadmap looks like. It's early on in the containerization strategy for this customer, but have they realized any business impact so far?

Hart Ripley (25:17.614)
They've definitely measured a few things and while I don't have all the numbers right now, there's been some occurrences recently where they've had some outages or some issues within the network. And the initial feeling and understanding and experience for them and for some of the executive team, the SRE teams has been that, well, we now actually know where the issues are now. We can trace them back much faster. That mean time to resolution that we kind of touched on a little bit earlier, they're able to kind of just dive in and say,

issues here, this is the device producing the problems, right? Because a lot of the times when you're dealing with network operations is you're getting a lot of noise and a lot of alarms from different devices, right? Different points in the chain. And the key part with this AIOps is you want to understand is, you know, is this router that has, you know, 10, 12 or more routers attached to it the problem, or is it something downstream? How can I hone in and look at the actual root cause? Right now they have an intelligent

operational network operational management system that says, we can give you pretty accurate ideas, visualizations, outputs of, yeah, there's lots of alarms firing, but here's the actual problem, right? Here's some data to support that and, you know, even suggestive ways to remediate those things, right? And I think we're starting to see that a lot with some of these other tools in the market now is it's the kind of the event -driven, the suggestive remediation where it's like,

well here's a bunch of things you can try. Right? And I think that can be helpful if you understand the context, if you have a plan in place to say how do we remediate? Right? A lot of organizations start off down the automation journey with well, you I just need to know where the issues are and then I can address them.

Hart Ripley (27:09.092)
Right. We still see a lot of large organizations not doing automated remediation because there's still risks there. The big thing for them is to determine where the problem is. We have a very capable team in some cases that can go and resolve that issue very easily. We have ways to do the resolution, but it's

biggest thing is actually finding out where the issue is without having too much downtime, having too much user experience issues, those types of things. that's kind of the biggest feedback we've seen so far is, now we have our sub ones reporting into our AI apps platform. So anytime there's an issue in the network, right, we know where it is, we can go to one spot and find it. We have one place to do notifications and alerts. And the remediation is much

simpler, quicker now. So that's kind of the biggest win today. There's some other integrations that, you know, I mentioned SNMP and stuff of that sort that, you know, can also plug into AI apps. Some of that stuff is still in progress. So.

You know, it's not all there today in terms of this client and their capabilities, but they have a platform that will support, you know, all these different types of logs and metadata getting into the platform. So the fact that they're already seeing some wins in terms of resolving issues and visibility, you know, obviously a great feedback to the business.

Marc LeBlanc (28:38.295)
would you say is next? you kind of laid out a bit of a path, but what's the next win or two wins they might be looking for?

Hart Ripley (28:46.887)
Yeah, so more integration into AIOps, right? But also building out kind of what is their SRE disaster recovery strategy look like, right? You know, we've got Watson AIOps in production.

But there's a lot more to say. If we have an event where our production site is down, where's our data? How has it been replicated to another site? How can we point all the sources and the feeds into this other site or into the storage that it's pulling it from and ensure that...

know, are not constantly able to handle real time issues, real time traffic, and basically flip back and forth between production and DR.

Right. So we've been focusing on how do we enable that? And it's not, it's not an easy task, right? It's, think a lot of organizations can do data replication pretty easily, right? Do backups pretty easily. But when you're talking about real time data and, know, especially events, you know, you're getting thousands of events a second. You can't really lose any data there or else, you know, you're not acting on kind of the most accurate data. Your deltas are off those types of things. So it's about how do we ensure that.

you know if there's an issue what the failover the business continuity plan looks like what's the actual process from an operational perspective to say okay we're gonna fail over here's the procedure here's how the data gets moved over here's the hot copies of the data all the feeds that are coming into AIOps you know

Hart Ripley (30:22.961)
Is that going now? Is that handled by DNS failover? Is that round robin? Right. So as we kind of touched on with building a platform, there's a lot of moving pieces and there's a lot of prerequisites and integrations. These are all things that we're making sure that, you know, we help them think about, but also take, go along with them on their journey to say.

These things are going to take some time. We need to test this stuff. We need to make sure that we have these documented procedures in place. But at the same time, getting them somewhere where they can have some business continuity when they need it. And that's kind of the big thing from the platform perspective. From features and functionality, they're going to continue to integrate more stuff into AI apps. There are some other features that they're not using today, like some of the automated remediation pieces that I was talking about earlier, the dashboarding stuff

still evolving. There's a lot of different network equipment, SNMP traps that some of the MIBs and stuff are still kind of needing to be updated or there's some older gear that may not support everything in terms of gathering as much information as needed.

I think anytime a telecom goes down this path is you have tons of different generations of hardware. Some of very old, some of very new, some of them with different capabilities than others. And you could have a great cohesive aggregating tool, but it's really dependent on how much data and what data can I get out of these endpoints to start with.

Continuing to spend it from a platform perspective, from a business continuity perspective, how do we manage multiple Kubernetes clusters, right? Supporting the application is a big thing. How do we do that centrally? To your point earlier, how do we do that in a kind of GitOps, DevOps fashion, declarative fashion? So, you know, as this organization continues to deploy clusters, how are we handling consistency?

Hart Ripley (32:19.723)
We have standards, we understand the hardware footprint in a lot of cases, but how are we deploying a cluster? How are we operationalizing the cluster?

We're continuing to layer in automation. We're continuing to layer in, know, GitOps and, you know, declarative outcomes and the way we kind of do the day two ops, right? As well as ongoing management, continuing to mature those types of things. But, you know, at the same time, we have to make sure that the team understands how this works, that they can do it themselves. You know, obviously there is some reliance on us as a managed service provider, but also making sure that, you know,

you

They understand what we're doing. You know, we don't want to be kind of that black box where we just make everything work. It's, you know, we see this as a relationship and a, you know, a good opportunity for us to work with the client and make sure that, you know, they're enabled, their individuals understand, okay, we've made an investment in OpenShift. We've made an investment in this IBM Cloud Pak for AIOps and other Cloud Paks. It's, you know, what are we actually getting, you know, in terms of outcomes? We're adding more clusters.

We're looking at other application use cases. Our infrastructure and other administrative teams now are understanding the benefit of containerization because we can move quickly. Our DR process looks a little bit different. If we want to add other tenants and what have you, we have the ability to do so. Another advantage of OpenShift, obviously, in light of some of the VMware stuff lately is we can now do virtualization on the same platform.

Hart Ripley (33:59.29)
same user experience, know, we have this client that's a VMware client and, you know, there's a lot of virtualization, you know, there. This gives them the ability to say, okay, what is our strategy around VMware? Right, maybe we wanna virtualize some stuff in OpenShift virtualization.

Maybe we want to keep some stuff in VMware. We're very comfortable managing VMware as there are a lot of organizations. So they've already got a platform that supports it. They're already getting familiar with the user experience of OpenShift.

Right. The, the automation capabilities, the consistency, the common user experience. And, you know, they can take some of that risk off of, you know, all our eggs are in the VMware basket. Let's say, let's look at some workloads or some lower environments, start getting familiar with Red Hat virtualization. And, you know, let's do a proof of concept. Let's start using our clusters for, you know, other use cases. Let's look at multi -tenancy and Kubernetes, right? These are all things that we've started talking about both head -on.

operate, but how to kind go down that path and educate these teams. And it's not just an investment to support this use case, it's an investment to support the business longer term, some of these critical systems, right? And then just give them all the really ultimate flexibility.

Marc LeBlanc (35:25.121)
Awesome. We always like to say that we touch on the people, process and technology in this podcast. And we definitely did that today. I think I've got three kind of main takeaways. The first one, by modernizing infrastructure with things like containerization and AI ops, you're really giving the network operations centers and IT teams a lot of our context around problems that are happening in the network. They can determine exactly what's going on.

where the issue is and they can reduce that time to respond a lot faster than if they weren't using tools like this and platforms like this. The second one would be around AI and ML tooling built into AOPs are really powerful when it comes to understanding the trends and issues on a network beyond just those KPIs alone. So it's giving the context as well. And the last one is really around that organizations shouldn't be afraid to use

partners as they're rolling in these new projects. It is a journey, as you touched on a number of times, not just on the implementation, but making sure that the right processes are there, making sure the right design is there, making sure that your teams have the right path to get up to the right capability. And leaning on strong partners that have expertise in those areas can really accelerate that journey for different organizations. Thank you so much to our guests today, Heret Ripley.

for joining us and letting some valuable insights into the conversation.

Hart Ripley (36:53.606)
Thanks for having me. It was pleasure to be here. Thanks, Mark.

Marc LeBlanc (36:58.593)
Thank you for listening to Solving for Change. If you enjoyed this episode, leave us a rating and a review on your favorite podcast service. Keep an eye out for our next episode, coming soon.

‍