Bank. We're delighted to have the CFO, Ken Miller, and the EVP of AI-Driven Enterprise, Sujai. You have to help me with your last name. Hajela? You're good at it. Is that okay? Okay. Let's just call him Sujai. He can, he can, we can call him just one word as business is accelerating to 63% growth. There you go. Anyway, we, very thank you so much for both of you being here. There's a lot of ground to cover, so just jump right in. Maybe just, you know, Juniper shares came out, and they're a little bit of pressure. Ken, I wanted to give you the floor to kinda maybe address that a little bit here in front of investors and explain, you know, what you've been hearing from investor conversations post the call and, kind of, you know, frame up the top kinda takes from the quarter. Sounds good. Thanks, Tom, it's great to be here. I will say we're likely to be making forward-looking comments in this conversation, and those do carry risks, so please take a look at our risk factors and our SEC filings. The quarter, I mean, I would, I would sum up Q2 as extremely strong quarter. Strong across the board. You know, enterprise, particular strength, you know, we saw 60%+ growth in Sujai's business, 38% growth in enterprise. It's now 45% of our company, and it's our fastest-growing vertical. That strategy is working to diversify towards the enterprise, but we still have significant differentiation and a lot of runway to go. Yep. We did see weakness in cloud, you know, more weakness than we expected, particularly on the cloud demand side, you know, new orders as they're digesting previously placed orders. To a lesser degree, we saw some weakness in service providers. That kind of demand weakness on, on cloud and, to a lesser degree, service provider, you know, impacts revenue, our future revenue. We did lower our revenue expectations for the year. That was obviously, you know, not well received, as, you know, no one likes a lowered, lower revenue number. Given the, the market that we see in cloud and SP, we thought it was the prudent thing to do. We're still focused on expanding operating margin. We're still focused on driving long-term sustainable growth and margin expansion. For the second half of the year, we are seeing some challenges in revenue due to the weakness in. Got you. ... in cloud, digestion in cloud, and to a lesser degree, service provider. That was really, kind of the message of the story. Extremely good results, a little disappointing, forward guidance based on cloud, cloud weakness. Great summary, and thank you for that. Just we're, you know, we're trying to get a feel from myriad companies that are reporting here on the macro. Just, looking out into the second half of this year, as you exited 2Q, you mentioned enterprise strength. Can you just kinda talk about what you're seeing, you know, looking out in terms of, you know, the year kind of started off kinda weak, you know, I remember end of 2022, the macro was getting soft, and just wanna kinda see where your, your head is at in terms of looking out to the second half from a macro perspective? On enterprise side? On the enterprise side. I mean, if you're talking macro, I would say the enterprise market is, you know, not particularly strong. I mean, I would argue it's, it's, it's weaker than, than the normal years, right? The macroeconomy is not providing tailwinds to the overall market. On the bookings perspective, there are still some backlog digestion, and you're starting to see, I think revenue is holding up a little better for some of, some of the market than bookings are. But you're starting to see a slowdown in demand, generally speaking. That said, that's a very much a macro comment. From a Juniper perspective, we're bucking the macro. I mean, that's the simplest way to say it. We are growing orders. We're, we're significantly growing revenue, and we're growing orders. We're definitely in take-share mode. We've said this before, we're 3%-4% of the market. Would it be nice if the market were robust and growing? Absolutely. That would help, you know, a high tide raises all boats, if you will. Is it necessary for us to grow and take share? Absolutely not, right? We've proven it. I think this year, where we're growing, you know, bookings and headwinds of a macroeconomy, 2020, massive headwinds in the economy, we grew our enterprise business. Mm ... the market, I would say, is, you know, weak-ish, but that doesn't necessarily give me any sort of concerns about our Juniper's opportunity within the enterprise. You know, first half, you've seen the results, you know, 30%+ first half. Second half, we expect, enterprise momentum to, to still be there and be our growth driver for us for, for the foreseeable future. Very clear. Switching over to Sujai, I think a lot of the driver of that enterprise growth is, has been your, your former company, Mist, that you're the founder and CEO of, now working here in, in Juniper. you know, just very succinctly, what, what's, what's driving this growth and specifically the sustainability of it? Yeah. I think what's really driving the growth is the outcomes we are able to deliver for customers. If you look at their very clear proof points, whether you talk about, you know, companies such as ServiceNow, whether you talk about, you know, campuses such as MIT, or you talk to the largest healthcare organizations in the world, they will clearly say that when they deploy the Juniper Mist network, they get an enriched end-user experience and a reduced network operation cost compared to a Cisco and Aruba. It's like you can go on LinkedIn, you can go on Google, you will see these screens from customers. What drives these proof points or these outcomes? It's fundamentally, you know, when we started Mist in 2014, 2015, timeframe, you know, there was a reason I left Cisco to start this company. It was, we felt the world needed a, an, a network which focuses on end-user experience. I always continue to say say it, because all current legacy architectures focus on whether the switch is working or not, whether the access point is working or not. We focused on whether you are getting a good experience on the network. This vision, you know, got really. Customers embraced it, and then the next question was, how do I deliver on this vision? That is where, and that also answers then the sustainability question. That is where we came out with this architecture, which is fundamentally microservices, cloud-based, and it is fundamentally AI-driven. We started this in 2015, 2016. I've given some TED Talks on this, and I can talk to you later if you wanna go into details. If you look at this architectural differentiation of a microservices cloud architecture and AI-driven support, the way we support our customers. You think about it, right? I recently heard, and I've been pretty public about it, I recently heard from a erstwhile competitor, which happens to start with a C, that they are now going to embrace a microservices architecture. Really? We started in 2016. That's the sustainability. Second is when you start going to the second part, which is about AI-driven. We are the only networking company in the world, on the planet, maybe in the galaxy, let's stay on the planet, right? Which uses our own AI to answer customers' questions or to support customers. No one else in the industry and networking does that, right? Cisco and Aruba haven't even started speaking about that. First, the vision which drives these outcomes was a clear differentiator, and sustainability is based on this architecture of microservices cloud and an AI-driven network. The sustainability is expanding the implementation. Do you land small and then expand? Oh, that- From there? Thank you so much for, you know, for pointing that out. I'll take an example of ServiceNow. I mean, they have been a pretty public customer of ours, they always take the example because they are leaders in IT service management, and if they don't understand that, who does? We landed with ServiceNow with Wi-Fi, okay? When they started seeing what Wi-Fi is driving for them, then they expanded, and we cross-sold switching. They saw the value of what they're driving with switching, and then we recently cross-sold SD-WAN. Not only am I landing and expanding a Wi-Fi, but I'm actually cross-selling Wi-Fi to wire to SD-WAN. You see this land and a cross-sell, and then within a Wi-Fi, or within switching, or within SD-WAN, we start selling subscriptions. It starts with basic networking subscription, adds an AI Virtual Network Assistant on top of it, adds Premium Analytics on top of it. Not only do I expand, then within each level of expand, I continue to upsell the value of the data which we create for that customer. That, it's like, vision resonates, technical superiority allows us the sustainability, and then the model supports this, you know, you land in $1, and you can start expanding it to $2, $3, $4. The company qualifies that as Mistified revenue. Yeah. That's not in the AI-Driven Enterprise, which accelerated to 63% revenue growth this quarter. So- Those follow-on sales. Mistified is within AIDE. AIDE. an SD-WAN or a... Yeah ... a wireline sale would be in that case? Correct. In that case. AIDE includes all wireless, which is all Mistified. 100% of wireless is Mistified. It includes all of our wired switching, LAN switching, and it also includes our SD-WAN, some security as well. The wired switching, SD-WAN, have been Mistified. Customers, that's where the growth is coming from. Yeah The customers that are interested in that part of our technology, but you can still buy a wired switch that is not cloud connected, that is not, leveraging the Mist cloud, but that's becoming the minority of, of that business. Yeah. The point which you're making, and, you know, Justin cannot hear, but I'll still say it. Mm. You know, if you look at what Mistification really is driving, we took it from Wi-Fi to switching to SD-WAN. I, you know, $1 of LAN gets me to $3-$4 of potential, and we started seeing this model, you know, manifested with customers. Here comes the fun part, right? The enterprise data center is there, right? Where we have the Apstra control point, which is doing amazingly well with intent-based. Now what we are looking at is how we... I've spoken about this before, and we are intently working on this, how we Mistify now the enterprise data center portfolio. The intent is driven by Apstra and AIOps, the value of network operations, which we have realized on the enterprise campus and branch, now extended to the data center. That's right. Well, Ken, back to you, with the harder questions in terms of the, the, the cloud. When we look out, you know, there's, there's been some concern about growth, maybe even into 2024, and I think you were very upfront, quite frankly, on the call about understanding that this could be multiple quarters in terms of cloud. Walk through, walk us through, like, the visibility there. You know, I'm sure you probably don't wanna- Mm ... do this again, so to speak, if you understand what I'm saying. Yeah. you know. That's a harder question. Well. Yeah ... you know, I'm not trying to be cryptic, but- Yeah ... you know, how, how much, so how much has been taken out? Then also, what's the visibility in terms of, you know, when you say something, this might come back in three to four quarters, that could be a little shattering to investors. Yeah. I mean, I have to admit, this is unprecedented times, right? Mm. We've been in unprecedented times for the last couple years. The first half of that journey was on the positive side, where orders were just going- That's couple. You know, through the roof, 100% growth in certain quarters in our cloud orders business, because of the supply chain constraints and the lead times and the early ordering required. 2 years in a row of, you know, 13%-14% growth in our cloud business. That's, that was the revenue side. That was what we were able to ship. Bookings was way north of that. Now we're coming down the other side, where they've already placed these early orders, and now they're burning the backlog. Backlog is coming down to normal. We're able to ship products within 30 days, 45 days in many cases, versus 12 months, right? It's a, you know, we're coming down the other side of this, which is a, obviously a huge headwind to bookings. Backlog's coming down. Revenue's holding up a little better than bookings, but still under pressure because of some of their own macro conditions. You know, the cloud customers are being more fiscally prudent than they used to be. I mean, that's the simplest way to say it. They're slowing down spend. They're doing some layoffs. You know, it's impacting their, their deployment timelines, which is actually pushing our backlog out a little longer than we expected. Okay ... which is putting some pressure on revenue. When it comes back, it's tough to say because I would've said six months ago, I wasn't expecting the, the steepness, right? I was expecting more of a gradual backlog burn, a little more gradual than it is. We knew it was gonna come down but didn't think it was gonna come down this quickly. Now that it's coming down quickly. The opportunity for bookings to resume, I think, is very high. I mean, next year, I fully expect cloud bookings to recover, and they're going to recover off a pretty low comp, right? This year's booking was, was artificially low because they're living off of 2022 bookings. Next year's bookings, there's no more backlog for them to consume. They're going to have to buy new gear when they need it. I think bookings will recover. The question is, does that bookings recovery, is it enough to generate revenue growth and offset some of the backlog headwinds we're seeing this year? The timing of recovery is uncertain. So- Is it early next year? Is it middle? Is it late? I'm not really sure. I think it will happen at some point. I know it will happen at some point. My gut tells me it'll happen at some point in 2022. If it happens later in the year, is it enough to offset some of the revenue headwinds of this year? That's going to be something that we, we just don't know yet. You mentioned in something interesting there about that maybe, the backlog has been burned off in cloud. Is that accurate? Absolutely. I mean, it's slowing down in all verticals, but in cloud in particular. I expect to enter next year with our cloud backlog to be at kind of normal levels, right? Mm-hmm. What you would expect it to be. That basically means when a customer wants to build something out in May, they're going to have to put their order in in the 1st quarter, right? They're no longer going to be able to live off a previously placed order. Just the booking order recovery will happen once they've drained through the backlog. The real exciting stuff happens when they build out, you know, continue to develop 400G, start thinking 800G, when they start thinking about AI clusters and the impact of that to their business. The long-term health of cloud, I don't believe is a concern of ours. We just need to get through this, this digestion period and recover, return to growth. I think that's a positive if that, you know, because you're still exiting the year at at least 2x... Mm-hmm ... backlog, and that's, so that's all enterprise and service provider. Yep. Sujai, maybe back to you on, you mentioned subscriptions being sold along these kind of nice money multiplier effects that you're getting to drive that dynamic growth. I know it's still early. I've asked you this many times, but we are getting probably into your fourth year, I, I believe. Mm-hmm ... at Juniper. Mm-hmm. Selling these contracts. What is, you know, and it's interesting to ask Juniper this for the, for the audience, a software question, but what is net retention rate, Net dollar retention rate, looking for those few or at least, you know, dozen or so contracts that are renewing now? Explain why they do expand if they are. Sure. First of all, you know, I wouldn't give a number, but it's definitely more than 12 or so. That's when I say when the model's panning out. Now, compared to last year, I can say, "Hey, it's happening. You have some, you have some, insight. Yes. Now comes the next part. You know, if you look at the net dollar retention and so on and so forth, I mean, it's a key metric, but, you know, you, you're right. It's a surprise you're asking this question of Juniper. Fundamentally, the way I look at it is we have SaaSified networking. You know, in our industry, at a macro level, Meraki started that first in 2012, now Juniper is totally SaaSifying the networking space. Our net dollar retention, it may be surprising for some of you, is easily north of 100%. It's easily north of 110%. It's easily north of 120%, right? My take on... This is based on our existing business, which we- This is just the software subscription. Just... Exactly. Fundamentally, if you say, "You know what? This is, kind of SaaS companies would like to boast north of 1.2 or 120%." Why? It's just because of, first of all, the upsell. When I sell, you know, let's go and say, like, I went and sold Jess a wireless network. He likes it. He starts seeing the reduction in network operations, but now he wants to take advantage of self-serving in that network. Hence, he has to buy the Marvis Virtual Network Assistant. Now, as soon as he buys that Marvis Virtual Network Assistant, that's a subscription. Guess what? You know, large language models are becoming a rage, you know, for whatever reasons. I guess, you know, ChatGPT suddenly made people realize, "Oh, AI is now right in my palm." If you really think about it, right, when we were working on chatbots, when we introduced chatbots in 2018 timeframe at Juniper, which was an additional subscription, we had started working on something called natural language understanding. You've got natural language generation and understanding. For us, the large language models have been awesome because we were the first one in the industry to announce an integration. ChatGPT is just a manifestation of an LLM, but we already announced that integration. That's yet another subscription. We announced Cloud NAC, which is about, like, okay, I've got the network, it's working, and for the cloud to get authenticated, you've got these two products out there, Cisco ISE and Aruba ClearPass. You know what? Vintage cars are good. Vintage network technologies don't work. Both of these are over 15-18 years old. What we did is the way we Mistified the network, now we are Mistifying the adjacencies around the network. We announced Cloud NAC. Guess what? Yet another subscription, right? First, within, like, when I sell you Wi-Fi, I go on increasing the value of that sale. The customer sees the advantage of what I bring with Wi-Fi, and they say, "Hey, but I have a switch connecting to the access point." We have a Mistified switch for you. Now you cross-sell, and then you garner the value of that. When I've sold a switch, "Hey, I've got SD-WAN, too." "Okay, let's talk about it." That's how it's... You know, I, I should probably do a visual of this, but I continue to go up the Y-axis within Wi-Fi, wired, or SD-WAN. I continue to go along the X-axis on the expand by cross-selling more. Right. I think that's the reason you see these NDR ratios, which I spoke about. Jess, Jess will have to put some software slides together in terms of, you know, number of subscriptions uptaked and, and whatnot in the next couple of years. First of all, we want to stop here with about five or six minutes left. Are there any questions from the audience? No. All right, let's keep going on. Maybe there were some interesting comments about the channel and deal registrations. Yeah. Just wanna maybe double-click on that. What-- I think Juniper's obviously there at Cisco in terms of using the channel. What is, what is your maybe if there's something unique that you're gonna go to the market with in terms of, invigorating the channel to go Juniper's way? I'd like to just get an update there and your plans there. Yeah, I mean, we've had great success there, and I think what's unique, quite honestly, is our solution, right? It's the, it's particularly the Mistification portfolio. It's the, it's, you know, the cloud delivery, AI-driven portfolio that, that Sujai is talking about. Quite honestly, what's, you know, the channel partners are taking notice. They're seeing that we're winning in the market. They see that we are the leader amongst leaders in the Magic Quadrant. We are taking market share, we are, we are growing quickly, and channel partners are coming to us, rather than us trying to push our solution that looked and felt a lot like the competitor solution. It's a differentiated solution, and that's what excites the channel partners. You know, Sujai mentioned a lot about the expand opportunity. It's about land and expand, right? We talked about new logos, up 30% year-over-year, last quarter, a record number of new logos. We talked about deal registration, which a partner generated demand, which is up 40% year-over-year. That's really the partners play a key role in that new logo capture, and then once we break in, we have the expand opportunity that Sujai mentioned, cross-portfolio, upstack, and software. That's really, the market, we're small, it's a big market, significant differentiation, land and expand, which is a par-- our portfolio, and then our go-to-market investment, adding more coverage, including channel, is really the factors that I believe give a sustained opportunity within the enterprise. You mentioned Mistification being a driver there. Does the channel get a cut of the subscriptions? Yeah, they get a percentage of the upfront sale. Yeah. Yeah. but nothing recurring. On the renewals, they get too. Just the renewals. They, they, they do not have they don't have a, an ARR business per se, but they have a second bite of the apple in year four and/or when we upsell licenses. That would get the channel very excited. Yeah. We're, we're also trying to ask, and this is that macro question, we're trying to ask, so to Sujai and Ken, you know, what is the nearer term monetization of, of GenAI? I know there could be some eye rolls there in terms of talking about that. I think we have to, though, given the groundswell that's going on in the market. Rami, you mentioned on the call a little bit about, you know, Ethernet and maybe making inroads into the InfiniBand kind of area of the AI clusters. You know, just kind of give us the, the outlook that Juniper has in terms of being able to leverage this kind of big investing cycle that we're seeing across the market, quite frankly, this last 6 months in GenAI, and then maybe some timing would be helpful, too. Yeah. Is this a 2025, 2024? You know, what kind of timing event? Yeah. So as it relates to the GPU clusters and the build-outs in the data centers that are all, you know, AI-driven, we view a couple things. First of all, you know, I believe it's the next killer app, right? It's gonna drive traffic growth as these services that are basically generated by, by the AI technologies become more reality. It's more users, more data, more bandwidth, right? That's good for just packet networking, right? Juniper obviously will benefit from that, from our wide area networking footprint, just more traffic, more growth. That's kind of an immediate benefit for us. The bigger opportunity, which is probably out there, you know, I don't wanna put. I don't think it's a 2024, possibly it's 2025. This is really more the network that actually connects those AI clusters, which today is predominantly InfiniBand, having that move to more an Ethernet technology, which is something we believe will happen over time. Obviously, there are others that disagree. There's kind of a bit of a technology battle going on between the pros and cons of InfiniBand versus Ethernet. We're on the Ethernet side, obviously, as are, you know, many of our closest competitors, Cisco and Arista, to name a couple. We do believe Ethernet opportunity will become a reality within the AI clusters. Probably not in the next few quarters. That's why I don't think it's a 2024 event. It's something that we are looking at as another, you know, pretty significant opportunity within the cloud space. Not, not baking it into any short-term models, but definitely focused on it from an R&D perspective to make sure we can capture that opportunity when it, when it does become a reality. It's probably early, have you taken a swipe at sizing that TAM? The market has. I mean, the market, the market's looking at a, you know, $2 billion, roughly going to $8 billion over the next three-four years. The shift to in, in Ethernet, most people think will be the majority will be Ethernet when you get out, you know, 3, 4, or 5 years. That's just the connectivity of the clusters to the networks? Exactly. It's all the networking of the actual AI clusters. That includes inside the clusters? Inside the clusters, absolutely. I see. That's where the big money is. It's, you know, it's data center connectivity of the AI clusters... Got it. ... intra data center. big, big growth. right now, lion's share InfiniBand, many believe that it'll still be a piece of InfiniBand, but there's gonna be a shift within a massive growing market to more Ethernet. Anything from you, Sujai, maybe then? Yeah. I think this may be something you may not be thinking about directly, but if you look at it, when Ken spoke about this, the world's talking about AI clusters, that's the supply side, right? If you look at it, there's a demand side of it, too. We just say the customer is gonna be leveraging the cluster, but what is it that they're leveraging it for? That is where if you look at our AI-Driven Enterprise or our ability to use Marvis, because there I'm the consumer of this AI supply, which is getting generated. We seem to be playing on both sides of the equation, and that's, I don't think people look at it that way. I mean, I look at it that way, sitting down with Ken. Mm-hmm. Guess what? Now what's happening is I just recently, just last week, was with, you know, one of the Fortune 5 customers, and they are like, "You know, my CEO, everybody, the line of business is saying, 'Generative AI, Generative AI, Generative AI.' Sujai, okay, should I enable ChatGPT for my support staff?" I'm like: "No. What you have? You bought this network from me?" "Yes." "Okay, guess what? That ChatGPT is the front end so that the network expertise is simplified for your self-serving customers, including maybe imagine, you in your own company, you can ask, ChatGPT, 'Hey, what's happening to my network? I'm not able to connect.'" Guess what? This ChatGPT is actually creating another rising tide, even on the demand side. Mm-hmm. Now I don't have to explain to them why AI for the network. They've already made this thing: "Hey, I need to invest money in this. What do I need to do?" I'm manifesting the use cases on the demand side. I know it's a very different way to look at this model, and it's gonna evolve over time, but we feel we are playing on the supply side, which will be all upside because it's not in the model, and it helps generate more demand. Yeah ... on the demand side, which is where we were the first AI-driven network in the company. Well, and it drives the conversation, which is important... Absolutely ... for a market share gainer. Exactly. Absolutely. Well, thank you very much for your time. We're out of time and very insightful. Thank you so much. Thanks, Tom. Appreciate it. Thanks.