 Good morning, everybody. Always great to come out and see this crowd. It's interesting the size of the crowd has grown, but the depth of the undertaking gets better every year as we move from concept and PowerPoints to prototypes to production. So a lot of folks are going to talk today and in the next few days about the evolution of where things are headed, where things are as an industry. So what we wanted to do today for the AT&T portion of the program is lay out some specifics of how we think about software defined networking, the open source component of it, and what our explicit roadmap is designed to do. And so what I wanted to do first, well, thank you, our Pete, for the introduction there. It was very kind of you. We have a great team on which I rely. So what I want to do is just tell you how we think about the evolution of our network. We, in the mobile part of our business, as we roll out technologies, we would go from 2G to 3G to 4G. And what we realize is when we go to 5G, 5G will be the first time that something is born in a modern architecture. And there were always compromises with us at AT&T and how we did things. And so what we did is we wanted to go back and look and say, OK, well, if software defined networking is what we're about today, then what's next? And what is our network of the future? And what I laid out here is a very simple schematic for how we think about things. In what we call our domain one program, our legacy environment, what we used to do is single-vendor optimization. Frankly, we were just a good supply chain. We were good at understanding the field of use and how we were going to apply something. And our job was to specify and purchase it. No more. That was a very difficult task. We could feel like industry leaders, but our job was to specify. And then we went into this next wave of networking where I remember in 2009, we got into this multi-vendor optimization, and we called that program domain two. It's really our first modern kick at a network. Our role had to switch to being an integrator. And if we were going to do this effectively at AT&T, we had to build the skills to be able to integrate. Frankly, in hindsight, that looks very easy. But at the time, we were extremely nervous as to whether we had the competency to deal with all of the complexity of the service chaining, of virtualization, of figuring out how to move from that appliance world into the complexity of more of a systems world. And then we got into network two, where we had to shift from being an integration leader to being the architect. So in software-defined networking, and I don't need to tell anybody in this room where the state of that was five, six years ago, we had to take on the role of architect. We had to take the risk that came with being first, going into undefined territory. And then what we started to do was ensure that not only did we think about every element of the network and decompose our traditional network into network functions, there were approximately 300 network functions. We had to virtualize those, and then we had to move them into a multi-tenant environment. And so you know a little bit because we're very public in our internal metrics that we wanted to get software-defined networking to 5% of the network functions, we made six. Last year, we set a target for 30. We hit well over 30%. And this year, we said we would take that number to 55%. But when we get to 55%, there's no turning back. You're more software-defined than you're not software-defined. And when you hit that tipping point, you start to say, OK, now what? Now what do we do when we have a software-defined network, and how do we capitalize on it? And that's when we looked and said, OK, the future of our network is not just about the architecture, but once it's built, how do we leverage that? And that's where we came with what we call our network 3.0, which is a data-powered network. If we do this correctly and we decompose our network into their primary functions, and those functions are properly built, we have a vociferous capability to gather data out of our network to go into a loop to build automation and machine learning that allows our network to be the best that it's capable of, not only for how we build networks, but for how we authenticate customers onto the network, how we deal with threats on our network. And so all of those traditional services that used to sit outside of traditional networking now become part of networking. And so our role in this data-powered environment is shifts to being a differentiator. I mean, we really have to understand how to use our data, and that's where we came up with the platform concept of Indigo. Indigo is really taking a page out of the software folks, whether it's the Android folks naming after desserts. Our logo color is blue, and we've said, okay, we're gonna start to launch platforms in colors of blue. So Indigo is our first instantiation of a software-driven platform. And so if you think about things in moving from traditional networking to this new world, Indigo is an abstraction layer that sits on top of traditional networking. So we have software-defined networking. That's a capability in our network that will be true of everything we do. This year our targets to get to 55%. We're gonna evolve our raw network technology for access. So the next trick in wireless is 5G, but we're gonna evolve everything we do to 5G capabilities, 5G speed, 5G software-defined architectures. And then the third and important element is how do you leverage that data power to create data-powered everything? So what we're architecting today is the abstraction layer that sits on top of access technology, software-defined networking, and the data that flows around our network. Indigo is that abstraction platform. That Indigo will evolve and accelerate over time just like your phone's operating system did. So as you think about ONAP and the network operating system, Indigo is how that gets manifested into customer, into services, and performance. And so to talk about some of the specifics and to make some announcements today, I wanna call up AT&T's Chief Technology Officer, Andre Fooge, and Andre will come and take Indigo a layer deeper. All right, good morning. Great to see you all. So I can't think of a more exciting time to be part of networking. And I hope you're excited too. In fact, I talked to my 11-year-old daughter this morning. She said, what are you doing this morning? I'm giving a talk to a bunch of network folks. She said, that sounds pretty boring. I said, these are the folks that are gonna change the way you use your smartphone. And she says, that sounds pretty cool. So I can't tell you how excited I am about where we are in this place and time in terms of where networking is going. It's really networking's era. And really the role that many of you play, and I know many of you are in different parts of this ecosystem, will really not just impact the various industries we serve, but really impact the world. And you're really gonna be part of something special here. So this is more about taking network to a whole new, powerful, impactful level that we've never seen before. And at AT&T, we're really excited to be taking this journey with you. And if you think back, Martin talked about 10 years ago where he started on this journey. And for us at AT&T, the real trigger point for us was also 10 years ago. And that was the launch of the first iPhone. And in fact, when you look from 2007 on, how much our network, how much demand has hit our network, it has grown over 250,000% in the last 10 years. That's pretty staggering. And what we quickly realized was the approach we've been taking for many, many decades, a very hardware-centric approach just wasn't gonna cut it anymore. And we realized that shifting to software was our best bet to meet this accelerating demand. We also realized that if we're gonna become a software company, we really needed a new architecture. And we really needed a new platform. But when we looked around at that time, there really wasn't anything off the shelf that fit. So we decided, well, what do you do? If you can't find it around somewhere else, you go and build it yourself. And so we decided to build ECOP. Now ECOP isn't some e-commerce platform or some sort of executive compensation system. ECOP is really a model-driven network operating system for STN automation. It's highly modular, it's very scalable, and of course, it's reliable and secure, and we have it running in our own AT&T integrated cloud. We've had ECOP in production now for over 2 1⁄2 years. It represents about 8 1⁄2 million lines of code. It's supporting well over 100 different virtual network functions. At all layers of our network, serving all levels of customers. And to give you a couple examples, we have it running, operating, and supporting our mobility network. Where every voiceover LTE call is partly managed by ECOP. Where we have our mobile packet core, we have tens of millions of subscribers. Again, being supported by ECOP. We also have ECOP supporting our enterprise business customers with one of our products, our network on demand product called Flexware. Where our enterprise customers can basically pull virtual network functions onto their premises and manage them as well as manage their bandwidth dynamically through our network. So pretty powerful. We also have ECOP supporting how our optical transport network operates. So all the various wavelengths that we switch across the country are now part of SDN and ECOP control. In doing all this, you know, we talked a lot about it publicly, and we also noticed that other operators were very interested in what we were doing too. So last year, a little over a year ago, we released a whitepaper about ECOP. In fact, John Donovan stood on this very stage a year ago and put out a call to all of you in the industry to say, please download this and let us know what you think. And we got tremendous feedback. In fact, we had over 7,000 downloads of that whitepaper and we had a tremendous amount of interest, not just from other mobile operators, but really many different parts of the industry, very interested in what ECOP has to offer. So last summer, after all this feedback, we decided to make a pretty big step. And that was to open source ECOP. So we worked with the great folks here at the Linux Foundation to figure out a plan, how to roll this out, and earlier this year, we released open source ECOP. While we were doing this, we realized there was an even bigger opportunity. And the bigger opportunity was to really work with some other operators that had another project going on, the Open O project. And this was an opportunity here to really sit down and create a global standard, an opportunity here to really align the industry on a single consolidated effort. And so just recently here, we merged ECOP with Open O and we created a new name, ONAP, Open Network Automation Platform. Now ONAP we think is tremendous opportunity here. Currently, as I speak here, we have over seven global operators representing nearly two billion mobile subscribers across North America, Europe, and Asia. That's over one third of the global mobile subscriber market signed on to ONAP right now. I'm even more excited about a whole pipeline of additional operators and service providers that are lined up representing about another third that will be joining here in the coming months. So we're really excited to see that ONAP, and John talked about the tipping point, ONAP really will become the global standard for service providers to introduce and operate and manage SDN. So you can see there's a tremendous amount of momentum building here, but the ONAP community is more than just telcos and vendors. We also see many web and cloud companies and integrators very interested in joining as well. In fact, just announced Microsoft just came on board, so we're pretty excited about that also. So ONAP is really about many things, but what's most important I would say is it's really about working together as a common aligned community here to innovate and deploy faster. It's about working together to build new things and not the same things twice. So I've talked a lot about how important open software is, but guess what? Open hardware is just as important too. So let me talk a little bit about an announcement we just released this morning that we're very excited about. So this morning, we put out a press release where we announced a very successful trial, what we believe a first in the telecom industry, where we put together a set of white boxes consisting of multiple suppliers, disruptive suppliers, and a lot of open source capability carrying customer traffic from coast to coast. So what did we do here? We collaborated directly with three merchant silicon providers, Barefoot Networks, Broadcom, and Intel, and we configured a set of different white boxes with each of these. We worked with two open networking ODMs, Aegema and Edgecore, and we stitched this all together with a single common open network operating system from SnapRoute. And this common operating system works across all the boxes and really provides a tremendous amount of capability to take this much, much further. We did a lot of interesting things here that we've never done before. We programmed some of the packet pipeline processing with the open P4 language. We did this running WAN carrier features such as MPLS. We managed the white box network with telemetry into ONAP. And this amazing team of collaboration did all of this in less than three months. In fact, one of the merchant silicon providers provided us a chip right out of the silicon foundry back in December. And again, in less than three months, we had that chip running in production with live traffic on it. I've never seen in all my career us deliver something from chip to production that quickly. And so it's really excited about the potential of what these white boxes can deliver. So think about it. It wasn't just about a half a decade ago what it took to process six terabits per second of packets. You'd walk into one of our data centers or one of our central offices and you'd see basically half a dozen refrigerator sized racks of routers. Now processing six terabits per second can be done on one chip that fits on the palm of your hand installed into a white box, the size of a pizza box and all controlled by an open common network OS. This is really fantastic. And we think this is gonna open up a whole new opportunity of things to come. So why did we do this? This is more than just about lowering cost and achieving higher performance. Frankly, that's table stakes. This was really about removing barriers, removing layers, removing all that internal proprietary API stack that we lived with with these legacy IT systems. Now we can bypass all of that and go straight to ONEP. This is about pushing software programmability as far down into the switch as possible. The furthest we've ever seen ever. This is also about removing all that multi-vendor CLI, removing all the complexity around what it takes to test all these multi-vendor routing configurations that would take, frankly, months and months and months for us to test. Now we've completely sped up the process here, again, by this very open common approach. This is really removing the barriers of entry of innovation. One great example I wanna highlight here is the ease of use of utilizing P4s in-band network telemetry. And we were able to use this on top of this barefoot Tofino chip. And it was really remarkable what we were able to see. This gives us, for the first time, unprecedented fine-grain visibility on a per-packet level. If this were like a medical conference, I would be talking about how we're moving from X-rays to MRIs. That's how big a deal this is, okay? Think about being able to find the state of every packet in your network. This is no longer the world of trace route and relying on SNMP MIBS that you can only pull every five minutes. In fact, trace route, I just looked up, was actually invented 30 years ago this year. So pretty rudimentary. Still most commonly used tool out there, probably. This is about getting fine-grain telemetry visibility that allow us to do some amazing things on the control side. Think about this like what GPS did for the smartphone. All of a sudden, when GPS was enabled with the smartphone, all these amazing new things could happen. Think about navigation. Think about how you order your food or you order a car to come pick you up. This is the kind of capability and the new sort of opportunities and industries that we think this new visibility will bring to networking. And this is really important to us from the standpoint of 5G. 5G is more than just about delivering much faster speeds. It's also about delivering ultra low latency, highly reliable services. You think of all the common use cases you hear about self-driving cars, how are those gonna work? How is that car, how is your car gonna know about that dangerous pothole, two car lengths ahead of you? When another car just went through it, wouldn't you wanna have a highly reliable, ultra low latency network that will ensure that that packet of that information to get to your car to move you into safety gets through? Those are just one of many, many new use cases that are gonna happen by getting this level of visibility. So with that in closing here, I can't tell you how excited I am to see all of you be a part of this new evolution. Networking is really gonna change the world. It's more than just making SDN better. This is about connecting lives and creating new opportunities. And this is really about helping make life easier and happier around the world. So with that, I'm open to take a couple of questions. Thank you. Any questions? I can't see, so. Hello, I'm Rajesh. This is from the Academy. As we know from the Pareto principle, only 20% can success. So what things you expect the academia should look into for the future? From this perspective, what? You have given. Well, so that's a good question. So everything's about the 80-20 rule, right? True. And so everything should be about, what are the things that are most important? True. Let me tell you in a couple areas that we need help with that I think the 20% should focus on. When it comes to new features and capabilities for the network, such as with network virtual, network functions, we really want to see functions that can align on standards. You know that old saying, you know, we don't want to build for snowflakes. It's really about building for Lego blocks. We really want to make sure that whatever you deliver is standard. And we'd like to see as much of that as open, but we realize too that there's a proprietary aspect that people need to make a business out of. And so how do you, you know, how do you do that? I'd ask you to look at this own app platform and all the capability it has because it really can allow you to do both. It can allow you to deliver things in a standardized way very quickly, but also has the ability where you can sell your secret sauce as well. And so that's one big focus area I would say that 20% should be focused on. Great, thank you. Regarding the standards, it takes almost a decade to actually become a standard. So before that, there probably might be some things where we have to trial upon so that that can be left. Yeah, so I would just say, and I can take one more question here because I've got the hook here on the timer. Thank you. But just really quickly on that, open source really isn't about standards. It's really about getting into a community, working together to align the community and to make sure this contribution gets adopted. So I would tell you from a standards perspective, we really want to focus more on open source. It's a much faster vehicle. Certainly there needs to be standardized aspects about it, but it's really working in these much more faster agile open communities is the way to go. One last quick question and then I've got a. Hi, Rick Merritt from EETimes, thanks for your talk. Can you give us any sense of under the covers what you expect to see in terms of your adoption of these white box systems and new kind of merchant silicon for a more programmable silicon in the future and what you want to see in those systems and chips? Yeah, so as I mentioned, it's really about, we're looking for solutions that are open, that we can do new and interesting things as I discussed earlier. And the opportunity here is huge. If you think about how the cloud is evolving, how the cloud is pushing, frankly, gonna push back to the edge. And you think about service providers such as ourselves and the way we're positioned. We have here domestically 5,000 central offices. We have over 65,000 cell sites, cell towers. These are all opportunities here for all these routing and switching solutions. And the time is really right now as 5G is coming on board here. This is the opportunity to get this right and to take advantage of these new approaches. So I think it's a really exciting time. And I would tell you, stay tuned, there's a lot more interesting things that we all have to share later in the year. So with that, I'm gonna pass the mic here. Thank you.