All right, welcome, welcome, welcome. We're starting right on the dot because obviously we have a virtual event here as well. This is being streamed, so we'll be respectful of everybody's time. I'm Jens Eckels. I'm on the marketing and communications team here at JFrog. Wonderful to see you and welcome all of you to Napa. Some of you we know very well, some of you we're getting to know. Thank you for joining us. I'm taking the place today of Jeff Schreiner, who you may know, and we're wishing Jeff a speedy recovery. Jeff, if you're watching, best of wishes. We know that you're here with us in spirit. We have an exciting hour ahead of us. We're going to interview AT&T as a customer as well. You'll also have the opportunity to ask some questions of some of our JFrog executives, including our CEO, CFO, CTO, about some of the announcements. Maybe you heard today some of the more financial questions that we have with the company. Before we start, I will read everybody's favorite part, which is the Safe Harbor statement. If you'll bear with me for just a moment, we'll get through that as quick as possible and get to our content. During this investor hour, we may make statements related to our business that are forward-looking under the Federal Securities Law and are made pursuant to the Safe Harbor provisions of the Private Securities Litigation Reform Act of 1995, including statements related to our future financial performance and our outlook for Q3 and the full year of 2025. The words anticipate, believe, continue, estimate, expect, intend, will, and similar expressions are intended to identify forward-looking statements or similar indications of future expectations. You are cautioned not to place undue reliance on these forward-looking statements, which reflect our views only as of today and not as of any subsequent date. Please keep in mind that we are not obligating ourselves to revise or publicly release the results of any revision to these forward-looking statements in light of new information or future events. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations. For a discussion of material risks and other important factors that could affect our actual results, please refer to our Form 10-K for the year ended December 31st, 2024, and Form 10-Q for the quarter ending June 30th, 2025, which is available on the Investor Relations section of our website. Additionally, non-GAAP financial measures will be discussed today. These non-GAAP financial measures, which are used as measures of JFrog's performance, should be considered in addition to, not as substitution for, or in isolation from GAAP measures. Now, with that part taken care of, we'll set the tone for the rest of the hour with a quick video. The shift has arrived, and this new era where machines are building for machines. This new era where machines are building for machines requires a different kind of foundational platform. Enterprises demand a firm foundational platform as a system of record for their software supply chain. For that, they need to manage containers and software packs, but also growing workloads of AI and ML models to deliver an unstoppable, proven power grid for the future AI-fueled software renaissance. As a foundational infrastructure platform, JFrog has formed strategic partnerships with NVIDIA, GitHub, ServiceNow, Hugging Face, and more to unify the software supply chain and place JFrog at the intersection of classic software development, security, and AI, with the emphasis on the software packages, the binary code. This uniquely positions JFrog to become the model registry of choice for the enterprise. This directly maps JFrog to the strategy of a modern CIO: consolidating tools, making strategic fit-for-purpose cloud decisions, weighing AI investments, ensuring security, and providing smooth integrations across the ecosystem. With our developer roots and enterprise trust, it's no wonder the majority of the Fortune 100 are JFrog customers. Our platform continues to fuel strong growth across every measure. Now, let's dive into the numbers. During Q2 2025, JFrog delivered $127.2 million in revenue, a year-over-year increase of 23%. Cloud continues to be a key driver of expansion. In the second quarter of 2025, cloud revenue grew 45% year- over- year to $57.1 million. Cloud now represents 45% of total revenue. Our security core products, which include JFrog Curation, JFrog Advanced Security, and JFrog Runtime, are gaining momentum and have been critical to our largest enterprise win. At year-end 2024, our security offerings represented 3% of total revenue, 5% of annual recurring revenue, and 12% of remaining performance obligations. This does not include the thousands of JFrog customers who use JFrog Xray for software composition analysis. This momentum is reflected in our RPO, which reached $477 million as of June 30th, 2025, a 75% year-over-year increase. Our enterprise customer base continues to grow as our strategic sales efforts bear fruit. We now count 61 customers with annual recurring revenue above $1 million, representing 45% growth year- over- year. Our fourth quarter trailing net dollar retention rate as of Q2 2025 was 118%, and gross retention equaled 97%. Our focus on operational efficiency has allowed us to achieve 16.3% operating margin during the six months ending June 30th, 2025, while adding over 1,500 basis points since fiscal year ending 2022. Our trailing four-quarter free cash flow margin is 29%, again highlighting our strong focus on operational efficiency. For JFrog, the rule of 40 continues to be a core guiding principle as we drive responsible investments with a balance in growth and profitability. This has allowed us to achieve an impressive trailing four-quarter rule of 51 in Q2 2025. We are not stopping at our own innovation. We are maturing partnerships with NVIDIA and Microsoft GitHub. We are working with ServiceNow to streamline operations across the digital business and deliver trust and compliance across the software delivery cycles. As we look forward to machines creating for machines, they too need a common platform, a system of record, to enable their digital infrastructure to release software with trust. Our customers tell us our focus on DevOps, security, and MLOps aligns with their modern software demands, powering JFrog's consistently strong business. Our technology continues to lead, and innovation is growing as we provide trust and speed to our customers' pipelines. We continue to be fueled by cloud growth, rising demand for our unified security solution, and a clear value in the world of AI. As we expand our platform into the universe of compliance, agentic AI, and model security, it is no wonder that even native AI companies are choosing JFrog as their single source of truth. The world is powered by software. Software is powered by JFrog. It is time to liquefy it all. JFrog, reimagining the software supply chain in the era of AI. Great stuff, great stuff. Now, I'd like to welcome with that our CRO, Tali Notman, and John Nuttall, Director of Technology for AT&T, to the stage. There we go. All right. Now it's working. Thank you. Great to be with everyone here. Thank you for joining us, and thank you, John, for joining me. This is not the first time that John and I are sharing the same stage, but it's a different audience, right? For sure, it's an opportunity for us to share with you the journey of AT&T and JFrog in the past, I believe, two years. It's been right since we started. I think it's a great opportunity for you to start maybe sharing, you know, the journey, the moment it started with JFrog, started with actually your own obstacles, right, at AT&T, and your own pain points that actually drove the modernization of software supply chain management to kind of the top of the list of priority. If you can share with us, with the audience. Sure, absolutely. I think the first key point was visibility, right? We had to be able to better answer simple questions like, who's on a CI/CD pipeline? The senior leadership were asking, and we couldn't answer because we had so many combinations of tools in our business units. We needed to be better at security. Things like the log for shell vulnerability proved that we had security folks pulling data from our platform at the time and putting it into spreadsheets and access databases to try and do the analysis that the tools should have been doing, right? We had a level, or you know, we were talking about this at lunch, an order of magnitude increase in the level of scrutiny that everyone was facing, I think, in the industry around Sarbanes-Oxley evidence. On top of all that, our binary platform at the time was having availability issues, and we couldn't look forward. We were so busy trying to keep our head above water, and when that goes down, that creates big problems. That was, that's where we were. I'm happy that we were there for you because this was exactly the point where we met, and part of the consideration of migrating to the JFrog Platform had to do with these pain points. What were really the main factors for you to consider the move, mainly when you're focusing on the binary management aspects of moving to JFrog's platform? We knew we needed, the company, our AT&T and the whole telecommunications industry has just completely changed from where it was at the beginning of my career, right? We used to buy very expensive boxes from Lucent and Cisco and Ericsson and Nokia, and now it's all delivered as Docker containers that run on commodity Dell hardware. We are truly a software business, and I needed a binary management platform that allowed us to move from just something IT worried about to the whole company worries about and not be, you know, instead of treading water, be looking toward the future. We had fortunately run into JFrog at an industry conference in Atlanta, Devn exus, and started to ask some questions. The more we looked, the more we liked, and the timing was just perfect when we arrived in Tel Aviv in July of 2023, where my boss's boss lives. Yeah, I remember that even now when we are discussing, even yesterday when we had our meeting, we were discussing the factory at the core and the single source of record, right? Which is like kind of your solid, trusted, kind of core to move your plans moving forward. Absolutely. That's it. We've realized that hope of getting to where we're looking ahead, right? We're modernizing our SaaS controls this year, and JFrog is right at the core of that. Everything pivots around being able to sign our code now and store that provenance information in JFrog and be able to use that so the Sarbanes-Oxley auditors look at managed access and managed change, and we're the primary owners of the managed change controls. We let security worry about the managed access side. We've announced a strategy internally to modernize our managed change controls back in February to the CFO organization that signs off on SOCs, right? And to our compliance organization and security. The salsa capabilities of GitHub and JFrog were at the very core, the foundation of what was announced. When we see presentations like we just sat through an hour ago, I have 13 people here from my team, including my boss, and we're all looking at each other like, this is really great. The ServiceNow integration, the GitHub integration, it just, it's a home run for us. I think that one of the biggest steps that we took together was really the migration to the cloud, right? The JFrog SaaS solution. Do you want to share a few? Yeah, we wanted to, when I took over, I've been in this job for seven years now, as long as I've been in any one job. We had a bunch of system administrators, but our customers were developers, and we wanted to get more of a developer culture on the team. We knew from the beginning we needed to get our tooling into the cloud. It just took us a while to get there. Now we're going to be 100% cloud by the end of the month. I shouldn't say that. Even JFrog, we have some on-prem stuff. It's more of a hybrid arrangement. We've got a developer culture on my team, and that has been a huge win because now when someone out in our customer area, internal customer area, has a problem, I have people that understand why that's a problem and they want to fix it for them. As we all know, you were talking about the presentations, but I can share with the crowd here that AT&T was one of the customers that got access to this information even prior because of the level of urgency around the governance, for example. We were talking about security, but you told us, hey, it goes beyond, right? Just security, compliance, governance. Maybe if you want to share a little bit about that. Yeah, I mean, we look at those as two related but separate things, and we're working both, right? We're doing shift left and JFrog Curation and GitHub Advanced Security, JFrog Advanced Security. All those capabilities have put, I mean, we've been talking shift left for eight or nine years in the company, in the industry, but we're making it real now. I shared at the lunch table, we've closed out, I don't know if I'm supposed to say the number, but a lot of secrets on the order of over 10,000 secrets in the last several months. The reason we're able to suddenly do that is because we have tools that put those numbers where the developers are, and then we can say, we're going to put a deadline on this, and if you don't fix it, we're going to actually prevent you from merging your code. We've got senior leadership backing us to do that. That's a very powerful position, and we're able to see movement of the needle, right? We can actually see the needle moving where in the past, security was very good at dashboarding things and saying, okay, you're a red, you're yellow, you're green, but when are you actually going to help them close it, help the developers? Putting this stuff in front of them and the tools they're already using was huge. Beyond that, the compliance piece is, you know, the salsa is going to take it from where we have 122, we have 3,300 apps in GitHub. 122 of them are SOCs impacting. Those 122 apps in the past, it's been they present their compliance evidence 122 different ways. Not quite that bad. I mean, consistent, but still everybody did their own evidence, right? We need to centralize that in order to get those folks back on focusing on 5G and fiber, right? That's where we need to innovate. We don't need to innovate around tools. We don't need to innovate around how do you provide SOCs evidence. We need to be innovating around 5G and fiber. The tools doing this for us is huge. It's a game changer. Seriously speaking, the end of managing compliance and governance on a spreadsheet, is it? I mean, we had a call for several years that our company controller, Sabrina, called the most expensive call in the company. It was every Director and above in the CTO organization that was involved in SOCs, plus in the CFO in the Controller organization. It was because, you know, with 122 apps, there's always, every month, somebody that made a mistake in their evidence presenting. You have to go back and make sure that nothing bad happened. Nothing bad ever did happen. The effort of proving that is huge, right? It's a big impact. We're not at every week calls with that group anymore. They're down to, I think, monthly. The promise now of what we're about to deliver with our new salsa pipelines is revolutionary. It really is. Probably the last. I'm sorry, I just wanted to add, I mean, the joke is, you know, we'll have Christmas back. It's literally true. There's been probably three Christmases in a row where a number of us, all the way, including the CTO, including the Controller, were on calls over Christmas holiday, statusing SOCs evidence, right, to make sure that everything went through. Everything was just fine. It's the proof that's the problem. That's what we got to automate. You know, I heard it so many times that I feel like Santa Claus at some point might surprise you in the office. You really are. The very last question really is about the future, right? Where we are, where are we heading from here? Of course, AI, right? Everything around AI, what does it mean to you at AT&T? Our Chief Data Office is the business unit that really drives AI within AT&T. It's been great because we're such a large company. It's been a great partnership that I was able to drive kind of the classic CI/CD part of it. Jeff and the whole JFrog team went over and had the early conversations with our CDO group to say, here's the AI capabilities we have. Here's how to manage models. I didn't have to go sell that, which was a real benefit to me. When they were ready and they had decided, okay, this is something we can benefit from, they came back to me and said, all right, we want your partnership. Here's what we want from JFrog. It's been great. I think what really is just to recap what I experienced working with you and with other customers is that we don't know what tomorrow will bring, right? Before, at the start, when we first started, it was mainly around developer efficiency and speed. From there, oh, wait, security is also important. These days, we're all talking about how to control AI, right? I mean, let it run, but also have better trust and control. I hope that from this point on, whatever the future holds for us, we'll be able to take it together. Thank you again for being here with us today. Absolutely. Thank you. Great. Thank you, Tali and John. Very, very powerful. We'll move now into a Q&A timeframe and a quick bit of housekeeping. You will have the opportunity to ask questions of our executive panel. Just in the interest of time, and because it will get late into the day on the East Coast, we ask that you limit those questions to one per person, please, when that comes. With that, I'll introduce the stars of the show. We have JFrog CEO and Co-Founder Shlomi Ben Haim, JFrog CFO Ed Grabscheid, and JFrog CTO and Creator of Artifactory Yoav Landman. Shlomi, I think you wanted to start us off with a few words. Can you hear me? Okay, cool. First, thank you all for joining us here. This is not taken for granted. Usually, we meet once a quarter over Zoom, and having you all at the biggest celebration of the Frog, that's a privilege. I hope we didn't just share a lot of announcements of what we've been baking through the year, but also every piece of these LEGO pieces build a bigger structure for you on what we are after. We started the day with Yoav's announcement about the first agentic repository in the world, JFrog Artifactory. To remind everyone here and in the Zoom room, JFrog Artifactory was the first binary repository manager in the world. Started in open source, then grew to what it is today. Fly is the seeds and the ground for adapting agentic behavior into the JFrog Platform. We'll start small with small teams that natively adopt AI, and we will adopt the best practices into our platform. Second announcement was delivered by Yuval, sharing with you the AI Catalog. We promised that models will be treated as first-level citizens at JFrog since they are yet another binary. This is what we do. First thing that we've built is the full automation, not only around model security, but around the model 360 lifecycle. An AI Catalog is the first step into the world of automating security for models. The third announcement was JFrog Uptrust. I'm sure that you will hear more of our audience speaking about that because this is not a piece of innovation that kinds of land to be tested. Obviously, every product needs to be evaluated. This is a known pain and a major pain that every company suffers from. Most of the companies that you know in the world are still managing their application delivery with spreadsheets manually. That's with the external and internal evidence that are crossing the quality gates all the way to delivery, fully automated, and even more, better integrated with ServiceNow to provide a one-platform experience. It's big news for us. It's big news. The fourth announcement was by the Security CTO, Asaf Karas, telling you about the agentic remediation. It's not anymore good enough to find the vulnerabilities or malicious packages or non-CVs. It's also very important to use the capabilities of JFrog Advanced Security for contextual analysis. A, tell me if I need to be worried about it or not. B, if I am, I want to remediate faster than everybody else. Obviously, the other capabilities we share together with our partners, we are very excited about this swamp up. I'm sure that the future is aligned with our vision, as we called it a long time ago, a liquid software world. Great. Thank you, Shlomi. Now comes the time to take your questions. Just for the folks in the Zoom, I do ask that you also state your name and your company as well. Jason, I'll start with you just because you're right here in front. Thanks, Jason Ader with William Blair. I guess the way you framed the four announcements, Shlomi, is it right to think in that order in terms of impact on your business, or maybe just elaborate on monetization and how this could contribute to your revenue growth over time? Yeah. We had six announcements. I counted four now here because translative dependencies and the extension of developer tools. This is a feature supported now, a better capability in the product, but it's not a new product. To your question, Jason, this is not the order of impact. I think that what you saw first coming from Yoav's presentation, that's a foot in the door to a whole new world of how agents are taking over building, curating, delivering software. There is a lot to be learned there. We will take our baby steps because there is one thing that we knew when we started to walk on Fly more than a year ago. What we knew is that with 7,000 customers, the majority of them are big enterprise customers. You don't take chances or you don't do experiments on what agentic is and how software delivery looks like. We'll take it one step at a time. I assume that Fly will be adopted first by the community and next by our customers. Even better than that, we will adapt the best part of Fly into the JFrog Platform. This is an internal disruption and not only a new tool for small teams. I believe that Uptrust is there, and it's going to be GA ready starting September 30, 2026. I saw the line in our booth, how many customers are already asking about it. You heard what John from AT&T said about it. This is a known pain. People are waiting to solve this pain. I think that this will have an amazing impact on 2026 and beyond. The last thing is security. We keep saying consolidating, consolidating, consolidating. It's not just another capability or another feature that we are adding. It's also how easy it is to bundle it around one platform. If you speak with the CISOs in the room, they will tell you we have too many security tools on our software supply chain and we need to control that. If you want me to kind of bet on what is going to make a faster impact, I think Uptrust would be the first. Great. Next question. Koji, I'm standing closer to you, so. Koji Ikeda from Bank of America. Thanks so much for doing this. I've been coming to SwampUp for many years now, and I'd like to say this is probably one of the most exciting ones, and you could hear from the customers that we've spoken with. Congratulations on that. With all the announcements, it seems like, you know, clearly AI is going to be a big theme for you guys for many years to come. I'm going to ask a more grounded question. How do we keep our expectations grounded from what could come from AI? Meaning, what are the customers looking for today? What are the problems that they're trying to address? What is maybe holding them back from spending much, much more with you guys? Question number one. Question number two, maybe for Ed, you know, a lot of announcements today, a lot of new growth drivers. How do we think about how this could potentially play into the fiscal 2027 targets that you have out there now? I'll take the first question, and Ed obviously will take it from there. I think that first, what prevents customers from adopting AI at scale is exactly what we say on stage, what we said on stage. First of all, you don't know how much AI will cost you next year or the year after, when you train your model, where you take your model, and so on. Second thing, and this is something obviously that other companies might build a solution for cost predictability. The second thing is that AI is still not trusted. Yes, there is a lot of pressure. There is a lot of pressure to adopt AI and to run faster. AI is still not fully trusted. The third thing is AI is still not fully regulated. Just a few weeks back, there was a new executive order coming from the White House about what it means to adopt AI in the enterprise. Obviously, these kinds of things are giving us an amazing opportunity to build together with the community what will be or how will be the future of software supply chain. Now, why we are so positive about the steps we take, assume that you have JFrog Artifactory as your binary repository manager, and you manage your Python there with PyPI and your Docker there, and your Conda and your CRUN and all other packages. What are the chances that as an enterprise with 10,000, 50,000, whatever number of developers you have, what are the chances that you will just decide one morning that you take a different model registry for your models, although they are binaries? If we have that, we can actually put a full stop right here. Let's speak about how big the market of model registry is. I think you understand how big it is. Let's speak about all the platform extensions to just the registry, and you will understand why we are so excited about it and why we have to act now and to lead the curve. When the market will come mature from a regulation perspective, compliance perspective, security perspective, cost predictability, JFrog will be, as it is for the rest of the binaries, the vendor of choice. When we released the model in 2023, AI wasn't even something that we considered or contemplated. It certainly wasn't part of our targets that we included. Today, what I'm looking at is I've delivered on the growth factors. I've delivered on security, and we've broken through those budgets. I've looked at the go-to-market in terms of the enterprise and what we've delivered in the enterprise with 61 customers, 45% year-over-year growth. I also look at what we've done in the cloud, exceptional growth in the cloud. I've also expanded my operating margins from that point over 1,500 basis points. I'm delivering a free cash flow of 29%, 29% over the last four quarters. I'm very happy with what we've done in terms of executing against our targets. The world has changed. There's a lot of uncertainties. There's macro uncertainties. There's geopolitical uncertainties. There are uncertainties that are being faced with a shift in the technology. We'll have to see what happens. We are being responsible with the way that we invest. We are being responsible with the way that we're delivering the message. Let's see where we're at at the end of 2025, and we'll go from there. Next question. Thank you for taking the question, Sanjit Singh with Morgan Stanley. When I think about how you got to where you are today, I would isolate it to one factor. It's sort of how quickly your customers update their software at a sort of first principles basis. As we move into this next area, you were talking about technology shifts. As we move into the AI era, do we think that the pace and rapidity of how customers update their software, and now I'll replace software with models, does that accelerate and drive growth for you, or is it a different sort of equation altogether because maybe the models are fundamentally bigger? In that kind of fundamental P times Q equation, on the Q side, how do I think about the volume of software releases and how quickly customers will update their models to drive growth for JFrog? Maybe I will let Yoav start answering about the adoption from the technology perspective, and then I'll take it to the business adoption. In terms of the pace of releases, this is going exponentially because now think about it. You don't have just the developers that are creating code and creating code a lot faster. You have AI agents that are augmenting them. At the end of the day, the end result of that is many more updates and many more updates with the AI quality test that is being added. Uptrust, by the way, is another quality test that AI agents can use the same way that you would use in a release that was produced by a human being. Your readiness to production just keeps accelerating. To answer your question, this is happening today. Customers are releasing faster and releasing more software. Business-wise, something like two years ago, we started to sell the security packages on top of Xray. It was a fair question that we got from our customers, like, are you a security company now? By the way, we got it from all of you as well. Are you a security company? What's happening? Not only that, we claim that we are going to consolidate all the point solutions into one solution. We called it JFrog Advanced Security, then JFrog Curation as a firewall, and then runtime. Slowly, customers started to see the value of not only consolidating tools, but also protecting the primary asset, which is the binary or the image or the model or the package. Now, JFrog Security gained momentum that you see companies at the size of AT&T adopting it all the way from creation of the code to the delivery of the code. I believe that the next expansion must happen with the MLOps because there is no organization on the planet that will just avoid managed models. It's not going to happen. This is not a trend. This is a revolution that already started. We can provide you already with the acquisition of Qwak with this technology on-prem and cloud. Last, I think that compliance and DevGovOps is going to be a big domain by itself because it's coming from a different wallet. It might take time, but we are sure about this expansion, especially when we are staying focused and disciplined. Our asset, our expertise is binaries. Around binaries, JFrog is the best company in the world. I think that this is regarding the adoption. Regarding the business bundle, we make it very easy for our customers to take a leap forward with JFrog. First of all, we introduce them to new bundles of security now, so they don't have to go by features. They just take it as is. We also introduce them to some introductory prices of the new products. It will not be a fork to swallow, although for us, it's a positive thing. Both from the business side, the enablement side of the sales team, and the product, we'll see the adoption coming in the next years, especially with AI. It will be a bit slower. Great. Next question is to you, Mike. Thank you. Michael Cikos with Needham. Two questions on my side. The first on Fly, and really, it's more to the architecture. Yoav, if I think about the keynote, I think you had said that Fly, you expect a gradual integration into the platform. Shlomi, just now you had said that Fly, there might be, or you guys expect to internally adopt this. Is Fly currently built into the platform today, or is it standalone and going to be built in? Can you help us think on that? The second piece, AI Catalog, totally understand the merits there. We walked through discover, govern, serve. How do we think about the different monetization mechanisms for AI Catalog? What does that look like when that goes live? Thank you. I'll start by answering the Fly question. Fly is running on top of the JFrog Platform. Like Shlomi said, it's an internal disruption to bring a new developer experience to the JFrog Platform. We'll adopt the successful parts of it to the enterprise level user and developer experience in the JFrog Platform. Behind the scenes, you can expect the same level of scalability and stability and production readiness that you have from the JFrog Platform. By the way, it's cloud only. We didn't say that. I'm sorry, Mike, can you repeat the question about the AI Catalog? From a business perspective? Let's see. First of all, AI Catalog, as I answered to Sanjit, AI Catalog is going to be part of a bundle that comes with Curation. The amazing success that we see with Curation and the logic behind curating your models as well, and empower Curation with the knowledge coming from AI Catalog makes sense from a technology point of view and a business point of view. Basically, it's a bundle of Curation and AI Catalog together. It would be an add-on on top of Curation that would be a per-seat add-on for now until other models will step in if it will not be a per-seat. Currently, it's replacing a lot of other security tools, and we want our customers to have kind of an apple-to-apple comparison. Next question. Right behind you. Hey, Miller Jump from Truist Securities. Thank you for taking the question. Congrats on a great event. I guess just hearing all the different announcements today really made me think there's a lot of exciting growth vectors, but at the same time, I think it sounds like it requires a lot of different specific R&D talents to execute on the different growth vectors that you have. Can you just talk about the extent to which there are overlapping skill sets that you can leverage in that versus, and also maybe, you know, for Ed, where you're looking to buy versus build there? Thanks. Yeah. You're right. What's the point on just expanding in our territory? If binaries are moving between one side to another side of the organization, just changing hands, and this is our best expertise, there is no one managing binaries better than JFrog, then I need to introduce myself to a new persona. This new persona last year and the year before was the CISO, and now it's the compliance manager and the GRC. What's the advantage of a company that scales with the focused expertise that we get the internal champions speaking for us? They are introducing us, as they introduced us to the security owners, they are now introducing us to the MLOps engineers and to the AI team, and hopefully to the compliance teams with the Uptrust and so on. What's the advantage of it? We are solving a solution that is being consolidated under the CIO. It used to be a world of CISO and CIO and other silos. Now everything goes up to the CIO. We see also with our strategic sales team, enterprise team, and upmarket, we see more and more CIOs being involved in the process of buying JFrog. The second thing that happens, and I'm hoping that you saw it today, it's not just our customers that are speaking about the expansion. It's actually the number one players in their territory that are saying, okay, we want JFrog as a system of record. By bundling us with them, we also have more than a foot in the door. We are very optimistic about it. I'm very excited about expanding the addressable market and not just how much I'm taking from the same persona. To follow up with your question, Miller, when we think about the use of capital, we did the V2 acquisition, which got the foot in the door around the security platform. We're now starting to show success with the integration of JFrog ML, which we did very quickly. Six months after the acquisition, we were able to integrate into our platform JFrog ML. As we start to think forward and what's next in terms of the vector of growth, we introduced today DevGovOps. That's a technology that we'll have to look at and see, can we build that internally or is that something that we do on a buy basis? This is something that we'll consider and we'll always be responsible about the use of capital. It's a consideration as we look deep into what our knowledge basis is and how we consider using that capital. I mean, I didn't mention, but I'm sorry, just to add on top of it, we know that Uptrust is not for a small developer shop. You don't run a crazy compliance process when you're 100 developers. This is for the big guys. DevGovOps with compliance teams for the big guys. It will take time. Someone will have to solve it. Otherwise, it will stay the bottleneck of software delivery. Therefore, not only that we aim there from the technology perspective, from the partnership and the too integrated to fail perspective, we also allow that on Enterprise Plus only with a package of DevGovOps. Also from the business side, there is an alignment with the expectation of where we go. Maybe I'll add to that. Every customer, every enterprise customer we spoke with, they already have something like Uptrust custom built in-house because as an enterprise, you cannot release your software without any governance. Yeah, so this relates to what Shlomi said about it's for enterprises, but it's de facto for enterprises. Right now, it's only easier to get it from JFrog and keep everything alongside Artifacts, alongside your releases. Next question. Thanks for taking my question, Shrenik Kothari from Baird. Shlomi, you just mentioned about the number one players in MLOps seeing the demand. It was interesting to hear directly from NVIDIA on stage talking about them using the artifacts on-prem and essentially all managed by JFrog Artifactory. You have talked about the hybrid model being a strategic advantage. Can you talk a little bit about what this endorsement really means on the ground? I mean, they are the number one players by any means. How does your new products also tap into this hybrid model approach? Yeah. I'll start from the end. Every product that comes from JFrog, aside of Fly, is hybrid. This is part of our philosophy. We are not compromising on that. We never compromised before. We will not compromise in the near future, mainly because of the fact that this is what our customers are telling us. I saw a lot of great companies twisting the arms of their customers behind the back, telling them you are moving to the cloud, and they ended up backing off and giving them whatever they wanted. My team knows that from the ground up, when they build a product, they build it to be hybrid. Now what we see, and we call it out in the last earnings call, what we see is that the lack of visibility and the lack of predictability of your cost puts some CIOs, back to Miller's question, puts some CIOs in a stressful point. They don't know how much it will cost them. They are trying to minimize the uncertainty. One of these moves that we started to see, and this is why when you asked us about the conservative guidance of the cloud, we called it out back then as well, we start to see a fit to purpose. No one is saying we are not moving to the cloud. They say we will build it to what we need and not just betting on the cloud like it's 2010 again. I believe that we will see more and more companies, and especially, that's surprising, but especially the native AI company, the biggest AI companies in the world are on-prem. They are not even trying to build any kind of data centers with the public cloud. Their data centers are their data centers, and this is how they plan to scale. I believe that in the near future, you will see a hybrid world, and we see it as an advantage and a differentiator of JFrog. All right, next question, please. Andrew Sherman with TD Cohen. Thanks, guys. Shlomi, the ServiceNow integration, that's interesting. Seems to make sense to me. Just talk about how that came about. It sounds like you have some complimentary customers. What is that customer overlap? ServiceNow has a lot of customers that you might not have. Can that drive that new kind of like GitHub partnership is done o r just how else could this show up in the model for you? I believe that what we see in the enterprise is that the admin of ServiceNow is not the admin of JFrog. The users, those who are working with ServiceNow and managing the application delivery, are going to the ServiceNow UI, and they expect everything on the infrastructure to come in. When we showed Uptrust for the first time in Q1 of this year to our customers, a selected group of customers in a conference we call LEAP, when we showed it to them, they said, "This is amazing, but the people who need it will not come to JFrog Platform. They will stay on the ServiceNow." If you can build this integration and become the pipes that drive the oxygen to ServiceNow, this would be great. This is where we reach out to ServiceNow. They had the same question as you have. They asked to speak with these customers. Part of these customers are the world's biggest bank, biggest retail, healthcare. They told them, "This is what we need." What you saw on stage was the result of that. It is not necessarily the same persona. There is a potential of a go-to-market collaboration with ServiceNow in the future. We have to test the water first and kind of walk before we run. I can add to it. Many of the customers that we spoke with, they actually, like Shlomi said, they said, "Okay, we have ServiceNow. Most of the organizational applications are already managed in ServiceNow. We need your application level, release management level application to be reflected because we already have exactly the same entity in ServiceNow. More than that, once we have a change request that needs to go for compliance reasons through ServiceNow, we want the same change request to be reflected in the JFrog Platform as assigned evidence so that our compliance can also vet for that." This was a very complementary solution from the get-go. We have time for two more questions for some closing remarks. Hey guys, this is Damon Gavinov from Barclays. I think some of the issues that you are seeing are not directly tied to AI experimentation. Can you just help us understand the durability and sustainability of some of the consumption you're seeing? During Q1, we assumed that it was experimentation, but as we moved into Q2, we saw stabilization of that usage. It could still be experimentation. We have to convert that usage into a commitment, and we're working strategically and hard with our customers to convert that into a commitment. We'll still see, it's not that we've converted all of that usage yet into a commitment. We're pleased with the direction that we're going, and we hope to continue to see that usage. As you know, that usage can go away tomorrow. This is one reason why we don't include it into our guidance. Secondly, it's important, and this is why we built the model that we did from a unit economics perspective and from a benefit for the customer, as well as for the visibility of JFrog to convert that into an annual commitment. We'll see where we're at. We have a few more weeks left of the quarter, and we'll give you an update in October or November, actually. Yeah, thanks, Mark Cash from Raymond James. The new bundle packages for security MLOps, you know, these started at 50 developers. I was wondering how this could help drive conversion from those just using Xray today. If you could talk about these bundles, do they have a pricing uplift versus the previous security sales motion you're doing. Just to repeat the question, we don't have monitors on stage. I'm trying to hear what you're saying. What you ask is about the MLOps bundle? Yeah. We have two lines of subscription. Just like JFrog, our pricing model went through a transition that followed the technology, obviously. We had to make sure that what we give our customers is also something that they can explain to their managers and manager managers. Don't forget, we are coming from a background of selling a $3,000 license 10 years ago. Now it's millions of dollars, as you can see in the numbers that I cherished. It goes up the organization. MLOps and DevSecOps are only available for Enterprise X and Enterprise Plus customers. DevGovOps, which is Uptrust and some other announcements you saw today, is only available for Enterprise Plus. The bundles, we have a bundle for security and a bundle for MLOps. The bundles for security come with different products for unified security and ultimate security, and same thing for MLOps. When it makes sense, we bundle them together. For example, the AI Catalog with JFrog Curation. The new pricing page was updated this morning at 6:30 A.M. It's all very clear over there. There are some kind of new offerings that we will have to listen to our customers' feedback and adjust ourselves. This is something that, based on our research and based on our very close work with our customers, this is what makes sense for us now. The way we see it from the JFrog perspective, we have an opportunity not only to scale with the new addition of security and MLOps and so on, but to also trigger an upgrade from Pro X to Enterprise X and to Enterprise Plus. This is a double trigger that we hope to see in the upcoming quarters. Great. We have time to squeeze in one more over here before we close. Thank you. Shlomi, you mentioned the AI native customers earlier being on-premise. When you think about how they've been using JFrog Artifactory, can you give us maybe some perspective on how that will help you take Fly and AI Catalog to the enterprise customer base? Yes, that's a very good question. As you know, our cloud business is based on consumption, data storage, and data transfer. Obviously, if you host more models or transfer more models, then we have an upside. That's aside from the add-ons that you buy for security, MLOps, and everything around that. In the self-hosted, it's a bit different. Therefore, we are still adding the security capabilities and the compliance capabilities, the MLOps capabilities by seat. MLOps is also coming with another unit price, which is ML credits because of the GPU and the compute that they will use. That's bundled into a package so we would not kind of lose our customers' focus with too many units floating there. Self-hosted will go with the add-ons mainly and more capacity for JFrog Artifactory. If you host your models there, we shared, if you remember, in the first quarter, a native AI company that started with JFrog with a couple of hundreds of thousands of dollars. They doubled it in the second quarter. They are already growing. This is with JFrog Artifactory because of the capacity that they need to allow to the amount of packages that they are hosting. By project, by capacity, by server, this is the self-hosted on top of it, security by seat, MLOps with ML credit and seats, and now compliance attached to the security offering. We have three minutes left, and we're going to stop promptly at the half hour. Any closing remarks from you, Ed, Shlomi, Yoav? First of all, thank you for coming to this event. It's, as you can see, an oversubscribed event. To see everybody here in the room, I get pieces of you, as Shlomi said, either in Zoom or at conferences. It's nice to be here in the room together. We had a lot of content that we delivered today. We're excited about where we're heading in terms of the future of JFrog. We're doing this in a very responsible way. We build this brick- by- brick, or as we like to say, leap by leap. We'll continue to be responsible in the way that we do this. I'm excited about what this has, what it will be in terms of a long runway for JFrog. We will update you again in November. Super excited about the releases that we're making. Uptrust is definitely hitting a point in my humble view. I'm not objective, of course. It really brings a way of controlling security rather than just reflective security. This is what customers need. John from AT&T mentioned the dashboards. Customers don't need just additional dashboards. They need a way to control. They need a way to govern. We bring this experience to them with Uptrust. Really excited about that. Yes, thank you everyone for joining SwampUp. The amazing opportunity for us is not just to communicate what we've built here in this room, but to give you the freedom to speak with all the customers outside, with all the partners outside in so many social events that we have here in Napa in the next two days. I encourage you to do so and hear a non-biased opinion about what is it that we build. I know that usually our investors are expecting a very fast execution on the business model. The moment we just release the technology, it will take time, not only because of the adoption curve, but also because of the disruption that AI brings to the world. Rest assured that if there is an opportunity there, JFrog can be the number one vendor. I'm not planning to be number two or five in other territories, but in my territory, to be number one, this is something that we are taking very seriously and we are committed to what we promised you that we will execute. This is our strategy. May the Frog be with you. Thank you very much for joining us. Thank you. That concludes our in-person event for the moment and online. You are dismissed.