Guy Daniels,TelecomTV (00:11):
Hello, you are watching Telecom TV and our special program on how Boost Mobile is leveraging Rakuten Symphony solutions to enhance network operations, service assurance, and customer experiences. I'm Guy Daniels, and in today's discussion we will look deeper into service assurance management through the partnership of Boost Mobile and Rakuten and discover the lessons learned along the way. Well, I'm delighted to say that joining me on the program today are Dawood Shahdad, who is VP Wireless Core Engineering at Boost Mobile, and Raul Atri, who is president of the OSS Business Unit at Rakuten Symphony. Good to see you both. Thanks so much for taking part in the program today. Can you provide an overview of the partnership between Rakuten and Boost Mobile and doward? Can I first ask you what were the primary goals and motivations behind this collaboration?

Dawood Shahdad, Boost Mobile (01:16):
Sure. Yeah. So Rakuten has been a partner for us pretty much from day one of our network build journey. So we're now nearing four years together as partners and we're in the operating phase of our network and we heavily utilize the Observer platform provided by Rakuten. Okay. At a very high level, it's a centralized framework that allows us to manage pretty much all of our network performance, whether it's faults, whether it's KPIs, or whether it's performing correlations across multiple domains. Today, I can tell you 100% of my core network functions, if it's packet core IMS messaging, anything that I have feeds into Rakuten, OBF, the observability platform and allows me to monitor and manage the network using a single pan of glass. So why is this important? Simply put telecommunication networks today, they're not for people anymore. They're going to be for machines. They're going to be machines talking to each other very soon.

(02:17):
There'll be billions of devices connected to the network. These devices, the applications, the use cases, they're becoming more and more complex and the needs to slice and dice the network and its performance at the same time while keeping up with the SLAs, it's going to get more and more complicated. So it creates a real problem for us as network operators. We just cannot hire enough engineers to manage this growing complexity, especially we just can't pass on the cost to the consumers. The solution is to create a network that can manage itself, an autonomous network. So teaming up with Rakuten is really our first step in our journey towards building a fully autonomous network that can manage itself in the

Guy Daniels,TelecomTV (02:59):
Thanks Dawood and Rahul, can you add some additional background to this partnership?

Rahul Atri, Rakuten Symphony (03:06):
I always believe Guy Magic happens when the right vision meets with the right team. I think that's exactly what happened with Boost Mobile and us. We, when Boost Mobile was conceptualizing the network they were building, it was all about cloud. It was all about network running, being dynamic and autonomous. It's so easy to look on the paper when we talk about automation, but the pieces are so difficult to manage together. Cloud networks are complex. Cloud networks are multiple layer vertically and horizontally. What we got as an opportunity was to stitch all these things together so we were able to monitor right from cloud infrastructure application services altogether, and when we are talking about core network applications, we're talking about 60, 70 applications coming together to deliver some of these services. So it's important and very important to create a unified view so that you can actually monitor the network services customer experiences proactively and also make sense out of it. So this partnership been wonderful for us both in terms of the practicality and use cases to deliver, but also to mature our product portfolio to also take it to the public cloud. And with partnership like this, we are very proud that we're moving into the right direction as industry to create a real-time telemetry driven network.

Guy Daniels,TelecomTV (04:31):
So let's look at this in more detail. Rahul, can you elaborate on the cloud observability suite provided to Boost Mobile? What are the key components and functionalities of this solution?

Rahul Atri, Rakuten Symphony (04:46):
I mean, there are a lot of key functionalities coming across as solution Observability is a very abstracted term, but under the hood it is all about KPIs, telemetry, traces, logs, and a lot more in the end. Where we are heading towards is how to make dark knock really happen, how to observe the network even while we're sleeping and how to be proactive in terms of insight, action, river insights, and that's what we were able to do together. We have now more than 10 OEMs being monitored. We have over 60 applications being monitored in time, and it's all about near realtime telemetry. Anything which is coming out of those applications can be pulled into the cloud native platform and we are working towards improving that day in day out. It's also T because it's across multiple different data centers and zones, we are also able to provide much more insights even before the incidents can happen and also reactively to convert those hiccups and changes in the networks and also correlate them into incidents and also monitor them proactively. It's that how can you convert those insights into a correlated incident so that you don't look at multiple things, but you look at one incident, which can actually be the solution or the RCA for the problem.

Guy Daniels,TelecomTV (06:09):
Thank you. Rahul and Dawood, how are you adopting this and what do you see as the key functionalities?

Dawood Shahdad, Boost Mobile (06:17):
Yes, so let me give a little background of our network really quick. So I think it's a well-known fact, we've talked about this before. Our core network is deployed 100% on public cloud. We are deployed on AWS, and we are in fact the only operator in the world that 100% runs the core network owned entirely on its public cloud. So it naturally made sense for us at Boost Mobile to implement the observability framework from Rakuten on the AWS cloud, pretty much co-located with the core network functions. So the way we have deployed rakuten's observability framework, we've actually got it deployed in three of the four AWS regions that cover the US naturally all of our core network functions that sit within AWS in the same geo-redundant manner, whether the functions are coming from Nokia, Oracle, Avenir metrics, pretty much everything feeds the performance counters, alarms, logging, and all of the critical metrics that we need to get insights out of our network.

(07:12):
They feed right into the Rakuten platform. So that's where the magic that Raul talked about happens. The flexibility that Rakuten provides, I mean it's immense. My engineers are able to create custom dashboards, whether we are creating dashboards for specific network functions, whether we want to monitor specific services, there's voice multi emergency services or special scenarios such as traffic, ramp up, traffic migrations. The platform itself, it is really easy for us to use and very intuitive and allows us to keep track of the different parts of our network and pretty much joined the dots as they were happening distinctly. So again, we've extended this platform to different parts of the organization and created customer alerts. NOC is a major customer for us. They rely on these alerts pretty much real time. So we're not having engineers sit and wait for alerts. We are actually having alerts delivered by the system to our noc. We then take action based on this. So again, tremendous help from this platform.

Guy Daniels,TelecomTV (08:10):
Yeah, and Dawood this transition from vendor specific element management systems, EMS and proprietary monitoring tools to Rakuten's solution. How has it impacted your operations? What are the key benefits that you've seen?

Dawood Shahdad, Boost Mobile (08:23):
Yes, so let's talk about the migration itself. So I have been personally involved with the migration away from the vendor specific EMS systems that we initially had in our network for a very brief period. I think initially when we were building the network, we had a bit of a rush. We had some deadlines to meet that were pretty critical, but I can tell you we started off with the vendor specific EMS systems. The intent always was to go and converse towards a centralized framework. With Rakuten, the journey has been extremely rewarding. We've come a long way from having to manage our custom solutions that often required vendor support. We had to pull in vendors like Nokia, Mavin Air and others for even making sometimes the basic KPI formula changes, changes in the way the dashboard looks and feels changes in the way they display on the knock monitors and whatnot.

(09:15):
Our goal was to make sure that we are able to hand off monitoring of the network performance to our knock so we can focus on designing the solutions of the feature and not have to constantly be engineering and spend a vast majority of time on the tools itself. Our teams, they love the product. It's a critical component of our network today allows me to make changes to parts of the network. For example, if I have to include a second vendor for a specific network function, I don't have to retrain my teams, my support organizations, my noc, my customer face organizations about ai, AP got a new vendor in the network now and here's how they're reporting their KPIs. We're able to basically swap it under the hood report the same KPIs like for likes, and I'm not having to go through this massive recent process. So again, there are things like this that we're able to basically achieve only because of the centralized in the manner in which this platform designed in our efforts to essentially force all the vendors to comply and feed their data into this platform.

Guy Daniels,TelecomTV (10:17):
Now, we've referenced this single pane of glass approach already, but how does having a single pane of glass service assurance solution enhance Boost Mobile's ability to perform correlations across multiple networks and service layers? Rahul, maybe we could start with you. Can you provide some specific examples for us?

Rahul Atri, Rakuten Symphony (10:39):
Sure. I think like Dawood was mentioning, we actually worked on three basic principles. Integration has to be super simplified. I mean, we don't have to take weeks or months to do integrations whenever a new node or a new partner comes on board. So integration has to become super simplified and easy. The scalability has to be there. The platform has to scale and become resilient so that it doesn't choke when the network grows to a certain level. And then on top of it, the programmability and configuration for the end customer in this case, we don't have to be there all the time. The customer should be able to program, create their own dashboard, create their own insights, even create digital workflows in case required to reach out to certain close loop actions. So I think we focus on these right things and internally we talk about this a lot.

(11:25):
Dawood and team have been the best product managers of the platform and being super supportive in case of how to use, how to elaborate, how to create this. And I think, like I said, a lot of things get simplified and sorted if we have a common vision and desire to reach out to this same goal. And that's how I think the unified observability platforms are done. There's all the data available, all the information, all the things insights available, but it's so simplified to take the next step of building autonomous networks goes into the zone where every operator wants to be because simplified, you have a unified view, you have unified telemetry, you can correlate them, and now you just need to switch them with closed loop actions.

Guy Daniels,TelecomTV (12:06):
Indeed. Thanks Rahul and Dawood, can you add any further examples for us?

Dawood Shahdad, Boost Mobile (12:12):
Sure. Again, my background. I've worked pretty much core networks all my life with the big operators in the US across the world. So again, when it comes to telecom networks, so you could have a wide variety of problems that you could thrown at. You can, historically you are not good at predicting some of the things that get thrown at you, but tools like this make it better for us to go in that direction. So I'll give you simple examples, right? So we could be having cell phones losing their signal or for example, within the IMS space, the voice space sip editors four, xx, five XX, et cetera. I mean, they can be signs of complex issues within the network complex. When I say complex, there could be issues with the infrastructure. We could have problems with the routing. Changes could have gone bad, we could have power if we've lost and all of a sudden there's essentially an impairment of the network.

(13:05):
We could also have to deal with some buggy software that's been rolled out that you couldn't catch in the lap, right? Firewall issues. You could have a firewall that's erroneously starting to block actual internet traffic that our customers are trying to browse data and they get blocked. So to effectively manage these issues, we've got to have a comprehensive wheel of the entire network. That means bringing all of the monitoring data into a single place, reducing essentially duplicate and unnecessary alerts and trying to join the dots as soon as you can. At the end of the day, it is all about how do you converge and how do you get to the solution or pinpoint the problem in the network fast enough, that way you can make the correct action faster. Again, if this same data was going into the multiple EMS systems, we could definitely come up with in-house, complex business logic, create rules, correlate all this stuff.

(13:59):
It's neither cheap, it's not fast. By the time we're done implementing such complex rules, the technology stack would've moved on, may have gone from 5G to six G or whatnot, and you're stuck in the cycle where you're always trying to get these businesses perfected while the technology is moving and now you're already late. So what we have done essentially by using this single pane of glass, we've avoided this issue where using the rakuten's OBF platform, we're now not only relying on the vendor specific performance data that is being joined together, we're actually, as I speak today, we're in the middle of adding our Act two network probes into the observability platform. So what does that give us? That gives us a second source of truth. So in case the KPIs, there's a question about reliability all of a sudden something sound right now we have a second source of truth and we can actually correlate and figure out if it's a source that's gone bad. Again, these are the flexible things that the platform allows us to work towards. It's adding more robustness and resiliency not just to our network, but also the way we monitor our network and really enhancing our ability to manage the network.

Guy Daniels,TelecomTV (15:13):
So as I understand it, with this new solution, Boost Mobile can build advanced analytics and closed loop automation use cases. Then Dawood, can you discuss some of these innovative use cases you are currently exploring or planning to soon implement?

Dawood Shahdad, Boost Mobile (15:30):
Yeah, again, I think one thing I want to clarify, it's not always just good enough to come up with new ideas. You've got to be able to look at the practicality and put them into practice quickly. So we are very particular about not only being able to detect issues across different parts of network fast, but also implementing solutions before our customers experience a disruption in service. So instead of looking for a one size fits all solution, we prefer a flexible framework that allows us to adapt to our specific network needs. We want to be able to address network issues and automate solutions quickly with without relying on our side help to build custom solutions that get outdated. Again, that's where Rakuten Observability platform in the partnership that Rahul talked about, where we're not really, it's not a vendor and a customer relationship that we have. We are partners in this. He allows by teams to feed in the usability requirements and whatnot. We're thinking how do we make this better? So essentially the uptake on this platform is wider and broader and we can simplify the challenges that other networks today are facing because they've been built over a period of time. They've got multiple EMS systems, they've got complicated use cases, they just can't untangle all of that mess. We're able to start clean thanks to the Rakuten platform.

Guy Daniels,TelecomTV (16:48):
Right, and Rahul, let me put this use case question to you as well.

Rahul Atri, Rakuten Symphony (16:53):
I truly believe that one part of the problem is to collect all the data from different sources. The other part is to provide a flexibility to end user, to create their own antics and dashboards and probably insights even to create alerts on the runtime. Because as you know, when the networks are building up and they're getting into a zone where we are onboarding customers, a lot of things keep on moving. There might be change requests, there might be optimization, there might be other things which happens to different platform and every platform making some changes would impact the overall customer experience or the other nodes as well. Maybe adding a new node in the network while you scale out might impact some of the traffic patterns to change. When you onboard more customers, the network will behave differently. Having a simplified unified platform helps you to understand those patterns.

(17:43):
I'm very sure when we integrate those probes platforms, it'll also be integrated with the customer experience. So how your network is performing and how your customers are feeling or how your customers are feeling when your network is performing into certain way is a great use case to achieve. And I don't think if you don't simplify or unify the data sources and the Observative platform, those kinds of things can ever be done. It's always easy, like I was mentioning, that you can obviously build a layer where you can pull all the data, but then you're spending a lot of energy and time to do those things which are simplified to solve. You can actually log on to systems and create your logics. The no can spend time, and I'm very sure that will be in the day age very soon where every NOC engineer or every operations person would be a data science engineer or an AI engineer. So this platform I think is a wonderful choice for going into that journey, obviously step by step, but I think that's what we are achieving together.

Guy Daniels,TelecomTV (18:38):
Yeah, thanks. You say step by step, there's obviously a lot of collaboration already going on between the two parties here. Can I move forward then and ask how this solution can position Boost Mobile to expand into enterprise and slicing use cases for monitoring and assurance? Rahul, what potential opportunities do you see in these areas?

Rahul Atri, Rakuten Symphony (19:00):
I always believe that Slice has two components to deal with. One is activating the slice and obviously you need to see which areas can actually support Slice, which nodes are able to support slide, what network components are loaded or would be able to support Slice. The other part is when somebody is paying dedicated amount of money for quality of service assured, they would love to have that experience coming through. So if you have given somebody, let's say an end-to-end slice, you would need to monitor that slice across the network. And if you don't have a unified platform, how would you do that? You'll probably have to indicate much more systems and Slice is going to be a dynamic network configuration. We are talking about quality of service on demand for certain users who are mobile and probably doing a lot of handovers in the network. And if you have to talk about those dynamic situations and let alone be the mobile network, whether it is a drone flying over cities or it's a smart car driving over on the road, these decisions are very time critical and are end to end and I think we are laying a wonderful foundation for those kind of use cases in the future to be taken care of.

Guy Daniels,TelecomTV (20:05):
Good to hear Rahul, thanks very much for that. And Dawood, you mentioned significant savings being made in infrastructure and financial resources as a result of using Rakuten solution. Can you quantify these savings and explain how they've impacted your overall business strategy?

Dawood Shahdad, Boost Mobile (20:25):
So I won't be able to go into the exact specifics of the cost savings, but I can tell you they're extremely significant versus if I had to manage and essentially pay each vendor for their specific EMS systems that they were providing. It's not only allowing us to simplify our operations, which is a direct essential CapEx and opex savings, we're actually avoiding the cost of extra infrastructure software licensing with the alternate path. Again, the platform also allows us to be smart about the data and how we handle data. We can choose to store some of the data locally at the edge. We can also choose to send some of the essential information to centralize locations for analysis. Again, we've got, like I mentioned earlier, we've got a network which is deployed across three regions, AWS regions. We try not to send data east to west, there's a cost associated over there, but we are actually able to customize and control how we're actually spending on this stuff.

(21:23):
Again, extreme flexibility. The way this is deployed, Raul talked about in his previous answer, he talked about slicing in some of those use cases. When we are hosting the enterprise networks off the different platforms and whatnot, we're no longer looking at multiple solutions for observability for those. Essentially this platform is capable of being sliced, diced. That way we can keep the data secure custom and essentially serve each customer within their SLA and expose all of their metrics that we have contractually obligated to exposing to them. So huge part of this stuff and the alternate route is go with each vendor, pay each vendor, write each vendor a check, just features that's extremely, extremely expensive and I can tell you because I do sign those checks and so we're glad we're not having to do that on the monitoring aspect.

Guy Daniels,TelecomTV (22:16):
That's good to hear. And Dawood, as we've already learned today, Rakuten solution has enabled you to remove several vendor monitoring tools. Can I ask you how this consolidation has affected your network management and operational efficiency?

Dawood Shahdad, Boost Mobile (22:34):
Yes. I'll start off with the history. Again. Our network is relatively new. We've been around for about four years. When we started building, I relocated to Denver. I had to pretty much hire all of the engineers from scratch here, right? So they come from varying backgrounds, but think about the cost involved in training each of the engineers or on these independent systems. Think about the constant need to do software upgrades, the underlying infrastructure upgrades. There are some, and I won't go into the names, there are some vendor specific EMA systems that don't work necessarily with the cloud, so I'd have to go deploy them on-prem. There's a need to constantly stay up with the security upgrades, do feature announcements, and essentially align all of this data together and make it look and feel kind of similar so that way the teams that are consuming this data, they're not confused as to this looks very different, but it probably means the same thing.

(23:33):
So we're no longer having to deal with all that complexity. We don't have to train our knock on multiple tools, engineering knock or even our customer experience teams. We all rely on a common tool for our network performance. So once you log into the tool, there's different views that you have. If I log in as an engineer from the engineering team, I may be seeing views that are specific to me. If I'm an executive, I may have executive dashboard access, nothing else. If I'm a troubleshooting engineer for specific domain, I will have access to those pages. Again, we're able to customize the different views and essentially provide each one of our organizations as well as the users, including execs, the views that they want so they get the maximum benefit out of this stuff. So try doing this with multiple tools and yeah, good luck. Let me know how that goes.

Guy Daniels,TelecomTV (24:20):
So this is all really encouraging. So let me ask you about the future plans for this partnership. Rahul, are there any upcoming enhancements or new features in Rakuten solution that Boost Mobile can look forward to

Rahul Atri, Rakuten Symphony (24:35):
You not know Guy how it is to work without? Right, so I personally believe that this is just to start. There's a wonderful partnership and I would emphasize on the word partnership. We both see the same dream. I would see the platform as it is own. I see the network assurance to be my part of the job. It's actually real good partnership on the platform side, we're working on a lot of enhancements. We're working on a unified hotel based, open telemetry based agent system, which will be one 10th of the footprint currently. We're also working to enhance our ETL stack, which will become much more native to AI for search and semantic search vector DB kind of situations. We're also working on making sure that we are much more adaptive to public cloud and some of the things we do today hardened open source. We are also looking at what things are easier to integrate with the public cloud information system and that's how we talk about the partnership that we are enhancing each other's capabilities and seeing one goal in the end and maybe in the future. This will also become one of the foundational systems for ai. We are also working on foundation fine tunes of our models internally in the platform and I think it's a great way to look into future together.

Guy Daniels,TelecomTV (25:52):
You are always innovating Rahul well, a final question then. What are some of the key lessons learned from this implementation? Dawood? What advice would you give to other telecom operators who are considering a similar transition?

Dawood Shahdad, Boost Mobile (26:06):
That's a great question. I can tell you over the last couple of years especially, we've been talking to a lot of operators around the world. We can definitely tell they're all challenged because they didn't get to build the network fresh like we did from scratch. But again, some of the lessons are to get the most out of your modern technology, the companies need to take control of their own network management. What does that mean? Take control of your own data. Data is an asset. That's an unbelievable asset that you have. If you outsource this task and you try to diversify and have each vendor manage its own, you will never be able to get all the benefit out of it. So that's where we've built a system where essentially our own software engineers can monitor and manage the network. Obviously it's with a platform, a great platform that comes from our partner.

(27:02):
It is really important for us to learn from others who actually build similar systems. That's where Rakuten actually comes in handy. They've actually used this platform in their own network. Rakuten Mobile in Japan. They weren't using it exactly the way we were. The stack was slightly different, but we've actually overcome all of that stuff and I can pretty much confidently tell you we are getting great support from Rakuten. We're actually benefiting tremendously from this platform. And any operator that I talked to, they would love to get to this single pane of glass concept. But again, we're looking to make this more and more mature. The whole idea is we don't want to be stuck in managing products and platforms. The network is built, the exposure of the KPIs and data is done. Now we need to worry about use cases, we need to worry about slicing, we need to worry about monetizing the network.

(27:59):
We need to sell the network. If you can't do that, then there might as well, what's the point in having a network, right? So that's where solutions that like this, they may be the building blocks of your transition to AI and autonomous networks, but they are super critical. You cannot jump in and start talking big about AI and all the stuff that you're going to do if your fundamentals are not there. And I think that's where Rakuten has definitely been a great partner. We consider this to be definitely a step one in journey towards that phase.

Guy Daniels,TelecomTV (28:33):
Great. Well we must leave it there. Dawood and Rahul, thank you both for taking part in our discussion and sharing all your insights today. Well, that's all we have time for. Watch out for more in our series, telecom reinvented by Rakuten Symphony. For now though, thank you very much for watching this program and goodbye.