(upbeat music begins) - First of all, fantastic to be here. We have so many AI innovations that we are gonna show you today, products and technologies that are coming throughout this year, and we are really, really excited to show you that. (audience applauds) I wanted the title of this keynote to be "AI with Desai," but Bill said that's not a good idea. So if we value today, I will thank my forefathers and next year it'll be "AI with Desai." Okay, all right. So we are a platform company and we have kept our guiding principles and technical principles pretty simple. Every single product is built on one architecture and one data model. When you think about the platform, so what is the platform? So the first thing, we would say, is the ServiceNow cloud that runs 24/7, 365. So that's the cloud layer. And what we have done in last couple of years that you have now Cloud Choice, where you can run ServiceNow in Azure as well as in AWS. When it comes to the platform core, this is the core platform as a service layer, which has all the technologies on which our applications are built. But you can do the same if you want to build applications on ServiceNow. And the major change here is that generative AI models are now native in our platform on which many applications are built. So that's the one change from last year and it's available right now. And then finally, all of our products or applications are built and prioritized based on the stakeholders that we serve, for all of you, whether it's technology, whether it's your HR department, security department, customer service, or for your creators. So those are our workflow applications. Our roadmap is pretty simple. We still do releases every six months. These are major releases. I told you last year, I'm not gonna pronounce the second one, I just can't. So first one, Washington DC just came out in March and Yokohama would be coming out next March. And if you are thinking, hey, what happens after Z? At least you should be thinking, at least that's what we are thinking. We start with A again, so that I can be for India. Yes? All right. So outside of this every six months releases, we will release AI at a much frequent pace, sometimes every month, as we learn with all of you based on the feedback that you give us when you use NowAssist. Today's show is very simple. We are gonna show you the roadmap on our core platform, then we will go into service management, employee experience, customer service, as in end customer service. And finally, we will end with AI, and we have a special guest today who will share with us on where AI is going. Okay, so to start with platform core, I would like to welcome on stage somebody who still looks like this after 19 years. He actually wrote a lot of core platform code. He still works very hard for all of you. And employee number nine, longest tenured employee, who has been at every single Knowledge. Please welcome our chief technology officer, Pat Casey. - Thank you everybody. So true story, that picture is about 20 years old, and at the time it was taken, the whole company worked in one little building and I'm reasonably confident you could have picked up that building and plunked it on this stage with room to spare. The whole building, including the outside. And if you'd asked me back then if I'd ever have a chance to stand up in front of a room like this, talking to an audience like you about all the cool stuff we've done, I think I would've laughed at you. It was just too farfetched. I will say it is incredibly humbling to get the opportunity to do that. And it's all still exciting, just like it was those 20 years ago. So with that said, I'm gonna talk to you about what we do with our platform, what we've been up to, and we have three kind of big themes we focused on in the last year. One, scale. Just what it sounds like. We want to help you get more stuff done with your instance. Second is performance. When you are doing work in ServiceNow, we want it to be faster, and pleasant, and work better for you. And the third piece is security. We want to help you get a better security model in place for your system. On the scale front though, I wanna start with a really kind of a cool announcement. We have an entirely new backend for ServiceNow, something called RaptorDB. This is a new database we wrote which is optimized for both transactional workloads and analytic workloads. And it's specifically optimized for the kind of workloads that ServiceNow has, the kind of workloads that you have. And that means you can get more work done on your instance, and as a side effect it'll actually be a little bit faster too. But to kind of give you some numbers there, we did a bit of a test here, and what we did is we took a regular instance of ServiceNow and we threw load at it until it basically broke, and you can break anything. And we got about 35,000 transactions a minute through it, which is a lot. It's pretty good. We then changed one variable. We swapped out the backend database, we put RaptorDB in there, and we got 93,000 transactions a minute. About 3x the workload. That means you can get more work done on your instance and it also means you'll have more headroom on your instance. So especially if you have one of our bigger customers, you can feel confident to put bigger, higher performance workloads on the system. When you're doing that work though, I wanna make sure it runs well, that you have good performance. So we've also focused on the performance of your instance. And I'll start by something we all do that's really important. We wanna make your upgrades faster. And it doesn't matter if you're a small customer, a medium sized customer, a large customer, everybody's upgrade is gonna be faster. Starting with the Washington upgrade. The benefits are the largest to the largest customers, but some of those top end customers, we're talking about a seven or an eight hour upgrade dropping down to about two hours. Really big improvement for you. And you'll see this every time you upgrade. It's not just about the upgrades themselves, though. It's also, once you upgrade, we wanna make your work faster. So we did a bunch of work improving the performance of our workspaces, and every single workspace, and we got a lot of 'em, got at least a little bit faster. The average workspace, about 25% faster. And some of the high volume ones you see up here, even bigger benefits. So if you're a workspace user, when you upgrade to Xanadu, this is partly in Washington, but mostly in Xanadu, your workspaces will run quite a bit faster. This is noticeably different. It's not just about some specialized transactions though, we also invested in the core transaction processing engine. So if you're one of our really big customers, you probably track the transaction response time. It's in instance observer. It's kind of a core response metric in the instance. With our Washington upgrade, we're seeing anywhere from about a 20% to a 45% improvement in overall transaction response time. Varies by customer, but everybody we've seen make this upgrade has seen a significant benefit on the backend response time. And so will you. It's not just about transactions though, it's also about helping you secure your instance. And on that front, we've made a big investment on trying to make it easier for you to get the right security model in place for your instance. We've done three things. The first, we added a whole new kind of ACL, something called a deny-unless ACL. So you don't just have to list every possible positive condition where something can happen. You can now tell us exception cases where something should not be allowed. It makes it way easier to put your security model in place. Second, we extended the ACL layer to apply to queries. You can now use security to control who can run queries and which fields they're allowed to run queries upon. And the third thing is we've added new features to all of our table based ACLs where you can specify the specific rows and columns you want them to apply to. You have even more granular control. You put all that together, we have got improvements on the platform, it's gonna scale better for you, so you can do more work, it's gonna perform better for you. So it's gonna be faster when you run it. And it's gonna be easier for you to secure. So with that said, thank you very much. Please upgrade, and back to you CJ. - All right. So it was exactly six years ago at this conference we created friendship and partnership with a very special company. I'm typically very insecure about my friendships, so as our engineering teams have collaborated quite a bit, I still ask them that, "Are you still friends? And can you tell everybody that we are friends?" So we have a special guest who's gonna say that we are still working well together. Can we please play the video? - Thank you so much CJ, and thank you to you and Bill for the amazing partnership. It's fantastic to join you at Knowledge 24 along with so many of our customers and partners. We're in the midst of a massive platform shift that will fundamentally transform productivity for every individual, team, organization, and industry. At Microsoft, we are committed to helping our customers adapt and thrive in this new age of AI transformation. That's one reason our longstanding partnership with ServiceNow is even more important today. Over the years we have brought ServiceNow solutions and services together with the power of our entire tech stack from Azure to Microsoft teams. And today we are taking the next big step, combining the power of NowAssist and Microsoft Copilot to help our shared customers improve productivity, enhance collaboration, and streamline business processes. Whether they're working in the NowAssist chat window or across Microsoft 365. I can't wait to see what you do with these new capabilities. And of course there's so much more to come. Thank you again for the partnership. I look forward to all we will achieve together in the years ahead. - All right, so our engineering teams have been working really well together. We wanna invite and we're gonna show you some product. Please welcome on stage, somebody who actually wanted to be a movie director with somebody like me who wanted to be a comedian to actually show you the products in action between what our teams have been working on for the past few months between Microsoft Copilot and ServiceNowAssist, Srini Raghavan, 18 years vice president of product Microsoft Teams. Excellent. - Hey CJ, thanks. - Thank you, so Srini, let's show the demo. All right? - Okay, thank you, great to be here. So rather than talk about it, we'll just show what we've been working together over the last few months. So imagine a scenario where I've been out for a few days and I'm trying to catch up after I come back. So I fire up Microsoft Teams, and right within Teams you see Copilot. And I ask Copilot to catch me up and it just pulls out the most important and relevant highlights of what I missed. So one thing that if you notice, I noticed that it says my laptop has been crashing many a time. I need to submit some logs. So I'm wondering if this is a great time for me to be able to just get an upgrade. So I'm going to use the power of NowAssist right within Copilot. And if you see the plugin here, NowAssist is actually enabled here. So this is where the magic happens. So I'm gonna say, "NowAssist, what's the laptop replacement policy?" And this is live demo. So there is the very first time that we are showing this on stage. So using Microsoft Graph and NowAssist, it goes and scours through all the different KB articles, thousands of KB articles, to be able to pull the most relevant information that it can. Not only it tells me what the policy is, but if you notice it also tells me when my laptop was last issued, as well as that I'm due for an upgrade, which is fantastic. So I'm going to actually say right here, "NowAssist, help me order a new laptop." There is a typo, let's see if it works even with the typo, with the power of LLM and OpenAI. And as you can see it just picked up right there. It's gonna look for how do I not just order a new laptop, but hopefully give me an action card, where I can take an action right within Microsoft Teams and with the power of NowAssist. So right there, as you can see, it just pulled up an order card. I can start the order. Now watch this very carefully, otherwise you'll miss it if you blink. The minute I say start order, it's gonna hand off seamlessly from copilot to NowAssist, which is on the left hand rail. So I just say order a laptop, it switched to NowAssist, and then I have to answer a couple of questions. I see the Surface Laptop 5, it seems to be a very good option. I'm going to say select the laptop and answer just one other question. I think I'm gonna go for this 15 inch laptop. It pulled up my home address right from my employee profile, right there, which is accessible only for me, shipped to my home, and submit the order. And just like that, within three to four clicks, I just went from Microsoft Teams, Copilot, invoking NowAssist to be able to complete a transaction that would've taken a lot longer. So back to you CJ. Hopefully you saw the powerful. Woo. I know you cannot tell, but my heart was racing as we are trying to do this live demo, and I'm still shaking. So now think about ITSM agent persona. This is, for ITSM service manager, a big deal, that we can do deflection. So I come to my beautiful service desk dashboard and I want to create a data visualization on all the things that we are deflecting. Okay? So it'll say, okay, here is a new generated, oh wow, I can add to the dashboard. Now I see the self-solve rate has gone to 73% and I can create a presentation for my manager, so Bill in this case. Can send data directly, one second. Everything on the top row, and create PowerPoint. Come on, come on. All right, so what you saw here is that going from Teams to NowAssist, amazing engineering collaboration past few months, many people haven't slept. Thank you very much Srini, thank you very much Microsoft. - Thank you CJ, have a great conference. - So when it comes to innovations in service management, we have three simple things in mind. Number one is, how can we improve self-service deflection that you just saw right now, while enhancing the agent productivity? So that's number one. Making your digital services more resilient with change. And we have a lot of things coming throughout this year on change intelligence, and most importantly, reduce the backlog that all of you deal with on putting more workflows on ServiceNow platform. So to speak about some of the things with one of the customers who have been trying this for now a few months with NowAssist is an organization called Teleperformance. Teleperformance is headquartered in Paris. They provide customer service to the iconic brands in the world, many of yours, 170 countries, 300 languages, and also half a million employees who've been trying NowAssist, so please welcome on stage Greg Weiss, SVP of IT and Digital Transformation on the stage. - Hi CJ. Welcome. - Thanks for having me. - So Greg, your scale is huge, and when you think about Teleperformance at the highest level, given your role, what are some of the challenges you're solving for? - Absolutely, so at Teleperformance, you know we provide customer care and other digital business services that are people led and technology driven with the ultimate goal of providing exceptional customer experiences. Now in order to do this we need to ensure that our associates, when they do run into IT issues, get up and running as fast as possible. So we have 5,000 IT technicians that are supporting over half a million users worldwide, and you can imagine the level of complexities with the different IT services and technologies that we need to support. They can introduce all kinds of challenges when trying to achieve that goal. - Totally makes sense. So I remember the call from your team that you wanted to lean in on AI. Why ServiceNow AI and what you really wanted to do? - Absolutely, so you know, as part of our digital transformation journey, we wanted to really drive through the platform more transformation. So we deployed process mining, predictive intelligence, and generative AI was a natural next step in that process. And so we looked at the analysis solution, and we really chose it because it was domain specific and really focused in the services around ITSM and processes that we wanted to deliver. And so that's why we went with the analysis product, the time, the value, and what it could deliver for us. - You're one of our first customers, so thank you. And what are the results, some of the results you have seen so far? - Well, pretty impressive actually. We went live in December, and since then we've actually seen, from an end user standpoint, leveraging the generative Q&A on the employee center, we've actually seen increase in our deflection rates by 100%. That's pretty awesome. - We've also, from an agent perspective, from a technician perspective, we've been able to reduce, on average, about five to six minutes of their time by leveraging chat summarization and resolution notes. In the major incident management area, we've really seen a big improvement here. So on average we're seeing 25 to 30 minutes of time saved per major instant management, which equates about 75 hours of productivity per month saved for that major instant management process. - Those are phenomenal results, and I know we will deliver more for you as we enhance our functionality throughout this year. So given the scale, Greg, what's next? - Well, what's next? We're really excited about continuing the digital transformation through the Now platform, focusing on NowAssist, and especially where we can help, from an end user experience standpoint, within the virtual agent. We're focusing on automation of productivity as well as change in knowledge management. We're also looking at the creator assist. We're excited about this. We think this will really drive our throughput from a development standpoint and get us that time to value really improved. We're also, as part of our strategic partnership with ServiceNow and Microsoft, we're moving our instances to the Azure Cloud. This further aligns with our overall global technical architecture that we're trying to achieve. And we're now a proud implementation partner of ServiceNow. So we're gonna take all the years of experience that we have at Teleperformance in digital transformation and leverage technology and customer workflows and bring those to our current and future clients to help drive further digital transformation. Really keeping people at the center, people led and technology driven. - It's pretty awesome. I still remember you became a customer in '22, went live in early '23, and you're one of the first to roll out. Pat Casey tells me you still make the most LLM calls in our cloud, so thank you very much. - Absolutely, thank you so much, I appreciate it. Thank you. - All right, so while Amy Lokey, our chief experience officer, and Joe Davis walk to the stage, here is what we are gonna show you quickly is that you just saw what Teleperformance has achieved, but with the functionality that is coming this year, what we are trying to do is make it easy for you to reduce your backlog, digitize the processes, make it available via NowAssist, and then see how it really improves your MTTR and all other metrics that you care about. So behind the scenes one of the biggest complaints we get from many of you, "CJ, we are trying to figure out our hardware return policy, Covid is behind us, we want to reclaim software licenses." And some of us have been working on, okay what does that process look like?" So I have a very high tech guest who is gonna actually be coming on the stage. Can we please bring the whiteboard? All right, so the point here is once you have defined the process, which takes a lot of work based on your own processes, mechanics, and everything you're solving for, how can we make it super easy for you to create a workflow and do the process? So Joe is gonna show you that also live. So Joe, go ahead. - Thank you CJ. So we've added a new input type to NowAssist. You can take a picture. So I'll take a picture of this process on this whiteboard right here. Okay, I'm gonna upload it to this machine, I'm gonna give it a friendly name. Okay, if we jump to the workstation here, I'm gonna create a new playbook, and I'm in our new workflow studio. And I'm gonna select that image that we just took, and instead of having to go find a developer and wait hours, or days, or weeks for them to create a process out of this, with NowAssist, I can do this in seconds. So we'll go ahead and kick that process off, and we're identifying the image and converting that into workflow constructs, and we can do that immediately. You can see the process here. That helped me get started really quickly as you can see, but it doesn't end there. You can continue to use NowAssist to build out the individual steps. I'll show you a couple quick examples. You can use flow to build out a flow step, and so I can just use natural language. And in this case what I'm asking for is to do some database lookups and to send an email notification to the employee depending on the results. And I'm just using natural language. A lot of people can access the technology using natural language. You don't have to learn the individual constructs and primitives of the workflow system. You can see we created a workflow really quickly as well. So it's not just about low code, we also have access to the full coding environment on the platform as well. So we can also come in here, and we're very excited to build code edit and code refactor so that you can use NowAssist to also improve all of your existing code. And what you're seeing here is it's recommending code changes in green and the old code it's gonna delete in red. I'm gonna go ahead and accept that. You can see that what we're seeing here is this is gonna turbocharge development, and really everyone can use a whiteboard, everyone can use a camera, everyone can build a workflow on ServiceNow. All right, and I'm excited to share with you how everyone can benefit from this process. And before I do, I first wanna say thank you to Teleperformance. They've let us use their beautiful brand to do this demo. They're not quite yet using all of this functionality, but I'm sure they will soon. All right, so as an employee of Teleperformance, I go to my employee center at home, and recently I've been having some issues with my device and my presentation software keeps crashing, but I haven't had time to deal with it. I've been getting ready for a big presentation. So when I come in here and I open my assistant, which is branded "Isabella," I can see that it has proactively detected that I've been having these issues with my presentation software crashing, and it can be resolved simply with an operating system upgrade. So looks like I can schedule this at a good time. So let's go ahead and do that. I'm a little busy right now though, so. "I'm busy, do this later." Okay, let's see what Isabella can do. Great. So Isabella can actually look at my calendar, determine that I've got a break at noon. So that sounds like a good idea. Okay, and because of our integrations, Isabella can connect with my Outlook calendar and book a reminder for me. Now we've kicked off this process that also identifies I've got some software on my machine that I haven't used in a while, and that's costing my company a considerable amount of money. I don't need these, so I can go ahead and relinquish them. Okay, now Isabella makes another proactive suggestion. People in my role tend to use Microsoft Project, and that would be a great software for me to get as well, too. So I would like to go ahead and take this time to get Project as well. Great, so that's been super helpful, but I just realized I've got a call at noon. That time that I scheduled, it's not gonna work. I hope I can change that. Let's see. And effortlessly that meeting has been changed, and not only that, there's a proactive suggestion again, there's a great company event happening next week celebrating an important culture with my teammates, and it looks like I can go to that. I'd like to go ahead and register. Okay, that's it. Problem solved. Everything's gonna get taken care of for me at a time that's convenient. I'm stoked. - So when Amy did the upgrade you just saw, she also installed a new video conferencing app. And since we just launched this app, I actually had to spend a lot of time troubleshooting, and it was just a permission setting. So I know though this is only the first of many and our team is gonna get flooded with this exact same request. So how do I prevent this? First, let's resolve the ticket, so I can log the solution provided. With NowAssist, you can see that it generated the resolution notes for me, saving me time. But you know, I think we can push it a little further. I can ask NowAssist to generate a knowledge base article for me. It takes the entire content of the incident, summarizes it, and generates the article. There you go. Let me see. I can still make edits, but as you can see, it did the heavy lifting for me. So it would take at least 30 minutes, now done in seconds. And everyone is able to self-service, including the chat bot that you saw who can automatically help people like Amy. Employees are happy and I free up time to do actual complex cases. - Thank you very much, Victoria. So this was live code, we are really excited to be shipping throughout this year, which I'll share a little bit later on the timelines for it. Now next, moving to employee experience. HR departments, global business services are inundated with many, many requests. We want to create a simple engagement layer and help out with many self-service requests. So we want to always improve employee experience for you and of course HR or GBS agent productivity. We are also trying to enhance the talent acquisition process, including onboarding and development of the talent. And finally, in this hybrid work environment, how can we evolve the workplace so that it's more collaborative? To tell you what they have been doing this with our HRSDs, in our HR service delivery product. somebody who works for an iconic company, is head of talent and employee experience. Please welcome Jeannine Carlucci, who also just recently got promoted at JP Morgan Chase. Jeannine. - How are you? - I'm wonderful, thank you. So Jeannine, given the scope of your role, tell us what are some of the challenges you're trying to solve for JP and the scale? - Well let me just start with, first of all, I'm honored to be here. Thank you for having me. I get the great pleasure of leading 310,000 employees across 65 countries every single day, and I really try my hardest to treat them like customers. So hearing everything that's going on here is actually super impactful for us. And what I'm really focused on now is all the experience with how we bring people into the company, develop them, upskill them, and make them have great careers. We're also extending that to the everyday experiences, we're calling it, at JP Morgan Chase. So thinking of things around the buildings you work in, the technology you use, the sense of belonging you have, we're really trying to take a real end-to-end view and work on the inside totally differently, as I like to call it. So we take the boundaries away and focus on that employee as our customer. - Yeah, and Jeannine, I personally love, so I use Chase, your mobile app is great, and in our first meeting you said, we want to treat our employees like we treat our customers. So why ServiceNow? And what did you plan to accomplish with ServiceNow? - You know, I thought about this question a little bit because I should have done this sooner. I didn't get a chance to completely influence everyone at the right scale. But we went to the cloud first, and did a great job moving all of our 60 platforms to one. We then took a step back, zoomed out, and said, if we really wanna do this across technology, CAO, compliance, and HR, we needed a layer on top, and what we're calling is integration as the name of the game. And ServiceNow really started to do that. So we went after help first, and we have like 325,000 hours that employees are spending on looking for questions, trying to get answers, looking for different things. So really taking that zoom out and going forward with this was something that I thought was a really powerful step for us as a company. Lori, Jamie, Robin, they've all allowed us to do it, and it's been really pretty amazing to get started on this journey with you all. I still remember when you went live, you were excited and delighted. So what are some of the results you're seeing from an employee perspective? - You know, when when I get to step back, one of the things we talk about all the time is what get measured gets done, and cohesion is the word we're using as it relates to that. So we're trying to take friction out of the way we work, we're trying to make sure we're using the data strategy we have to fuel personalized experiences. Obviously ServiceNow is a key component to that. And as we think about the 80,000 people or so, 'cause it's plus or minus every year, that we onboard, and the reputation that we have as one of the greatest companies in the world, we really wanna focus a lot on the onboarding experience. Because if we can get you through the doors and you feel exactly like the brand of JP Morgan Chase, we can continue that journey with you as we keep going and move forward with you through the rest of those experiences and the life cycle of the way the moments that matter for you at JP Morgan. - That is pretty awesome. You know, when you told me 80 to 90,000 employees a year that you onboard, that's 7,000 ish per month, that's close to 2,000 a week, that's 400 a day. - Thank you for the scorecard. - Yes, thank you, and we want to help you improve that. But just give us on onboarding, that's the next thing you wanna solve with ServiceNow. What are some of, how many steps that an employee go through and where do you want to simplify? - Absolutely, so look, when we onboard employees into the JP Morgan Chase, I hate to say this, but we do have a lot of friction because of, like I mentioned, the lifecycle we go through. We've done a service design and looked at about 525 tasks, 18 different applications, 64 policies people have to read. So our zoom out with this has a real reimagined aspect to it that we're really looking forward to, because we get that chance to really take the workflow to a whole different level, and like I said, integrate across JP Morgan Chase so that that employee stays front and center as the customer. - That's pretty awesome. We are delighted to partner with you. Please thank Robin and the team, and thank you for being here. - Thank you. - Jeannine Carlucci, ladies and gentlemen, JP Morgan Chase. We also want to give a shout out to one of our awesome customers, that's Kellanova Brand. My son used to eat waffle every single day and Pop-Tart every Monday. So we have a special giveaway from our Kellanova team. Can we focus on our Kellanova team? Thank you very much for treating all of you. So if you see in front of you there will be some envelopes with some special treats from Kellanova brand. You can use it later on. For their successful employee experience go live. All right, so our next chapter is around end customer service. When it comes to end customer service, as in your customers. And when I say your customers, again, your end customers, not your employees. We have prioritized a few key things here. The first thing is just improve customer experience with our self service, like we saw with IT. We also want to make sure that we optimize the customer service agent experience for the contact center, and transform the lead to order process by providing middle office sales workflows, what we call sales and order management. So to speak about this, I'm delighted to welcome, from one of the most iconic 165 years in business, $40 billion in revenue, 30,000 employees, 15,000 agents who work for them. She has been hall of fame CIO, also top 50 women leader. From Travelers Insurance, please welcome Mojgan, EVP, CTO, and CEO. Mojgan, please welcome. - Hi, CJ. I like the gifts. I see your golf shirt clearly. So there is a message here. What would you like to share with us? - All right, well so let me explain. So as you said, travelers provides auto, home, and business insurance, but we're also known for our iconic red umbrella, and for the Travelers Championship, which is one of the top four signature events of the PGA. So if you look under your seats, some of you are gonna see an envelope that has a red ticket in it, and you can exchange that at the expo either for one of these or some of these golf balls that are also championship golf balls. And there are, these are yours by the way, CJ. Oh, okay. - And there's also four lucky winners who are gonna have a ticket for two people to this year's championship, which is June 17th through the 23rd in Cromwell, Connecticut. All right, so that's very generous of you. Mark, if you can help us out. That's very generous of you. Hopefully I get a ticket and I get a chance to be in Cromwell, Connecticut as well. So Mojgan, you have a very massive role. Can you just give us the scope of your role at Travelers? - Sure, so I say look, it's everything technology, from our backend claim and underwriting systems, to our customer solutions, and then everything operations. So customer service, billing, business resiliency. And this is for our 6 million customers. - That's a lot of customers. So tell us what problem you were trying to solve for claims, which is a core part, claims processing, in the moments that matter you are trying to solve and why you chose ServiceNow. - Look, at its core, delivering the Travelers promise is really being there for our customers at the most important moment, which is during a claim. And so at that moment our claim professionals are not only doing the technical parts of their job, but they've gotta be there with empathy, and provide clarity, and really reduce the duress that's on our customers. And so it's really important to us to make sure that we provide them with an environment that allows them to do that job effectively, and that's what we're in the process of doing, really creating those platforms for them. - Yeah, so because if I'm calling for a claim, then you want your claims professional to be servicing me rather than worrying about technology. So I know you evaluated many solutions, but why did the team choose ServiceNow? - So, you know, as we've been modernizing, having a modern architecture and you talked about this earlier, CJ, I heard you, has been critical. We wanna make sure, first of all, we wanna make sure we buy and partner if we don't have to build ourselves when viable, and we wanna make sure that everything is cloud native and really based on a modern architecture, data architecture, data model. And so when it comes to orchestrating the claim throughout its lifecycle, and really having visibility into the status of the claim, I'd say the ServiceNow platform was absolutely perfect, aligned very much to what we were looking for. - That's pretty awesome, because claims orchestration being core to Travelers when moments matter for your 6 million customers, we are excited to partner with you. I know the team has some impressive go live dates And we look forward to making sure you get value out of the platform. My last question is, so claims is one piece, what next? - So, and as you know, we're also doing a lot, with IT, ITSM, iTalk and so on. And so we're really in the process of continuing to serve that. I just also talked about some of the gen AI capabilities that you guys have claimed were on our way. We're also doing some proofs of concepts in underwriting and operations. Looks very promising, so we'll be talking. And of course, you know, as everybody else, we are exploring AI. We've got quite a few things that we ourselves, from a gen AI and large language model perspective, are in different parts of like development and rolling out. But we're also very keen to really leverage anything that you embed in the platform, CJ. You're a very humble person. CIO hall of Fame from 2022, thank you for being here. - Thank you for having me. - And Travelers for being a great customer, and thanks for your generosity. - Thank you very much. - Okay, so now about AI, and in this final chapter I just want to outline, first of all, what is our very simple strategy? Our simple strategy is intelligent workflows in the flow of work. That's it. We just wanna make the workflows that you are using, or you make, create, new processes work in the flow of work. That's our strategy. Now how will we execute on this strategy? The execution is also very simple. Last year I shared with you some domain specific models and a few other things. But our first thing is, we are going to create our own models. We already have. We're gonna create our own models, we are going to release them via NowAssist based on the use case. And then we will make sure that those intelligent workflows get executed for you once you use NowAssist. So that's our strategy. Create our models, make them enabled via NowAssist, so that you can have intelligent end-to-end execution. So let's talk about gen AI models. We absolutely love open source community. We take the best model that is available, we will fine tune it, and then most importantly, for ServiceNow use cases and ServiceNow data types, we will train these models so that you don't have to train this model. These are models that are just available for say case summarization, or incident resolution, or creating a KB article, or picking multiple KB articles and summarizing into one. So we will take the best of general purpose models combined with use case specific models and run that, as I showed earlier, in ServiceNow cloud. Through this massive open source community, we have a fantastic team at ServiceNow when it comes to AI, from research, science, applied teams, engineering teams, platform teams, and experience teams. So that's what we are gonna do. Why does that matter? When the models are very specific to our use cases, first of all, your data does not leave ServiceNow instance. Number two, they run faster because the models are smaller. Otherwise the large models still you will see a lot of latency versus you saw today how fast it is. The accuracy is high, and of course your data remains private. So that's why it matters. This is behind the scene. We continue to work, Joe Davis and the team continue to work behind the scenes to create these models. However, if you say, and some of you have said this to us, that we want to bring our own models, we want to use, you know we have done this, we have done that particular hyperscaler, totally fine. If you want to bring your own model, you can, and it can be either an "and" or an "or." So we currently support OpenAI, we support Microsoft, we support Google, Watsonx and others. But if you decide to do so, that you will have invested a lot in language models, ServiceNow will connect to the models that you have running within your premises. That's totally fine. But we believe that our use case specific models is a unique thing for your use cases that you use on ServiceNow. Now the second piece that I'll say on analysis roadmap, the teams are working really hard, it's a very aggressive roadmap and you see on the left, in the middle, and on the right, throughout these years you will see NowAssist being enabled for some of the key products that are in use by all of you. And what Joe Davis showed, I'm now gonna share some dates with you on generation roadmap. Our goal is to make sure that every single object in ServiceNow platform can be generated using either a text or an image. Okay, that's our goal. So multi-modal language models, or multi-modal models, however you wanna say it, we release code generation. Now as you saw Joe showed ability to edit it, you can generate workflow, you can generate spokes, integration spokes and so on. This is a very, very aggressive roadmap that we are gonna execute on and that's what we are really excited about, that you can just type text, or you really focus on process engineering or re-engineering and you don't really worry about writing the code. Now, intelligent workflows, so what do we mean by that? There are three aspects to intelligent workflows. First, because it's a knowledge conference, we are really excited to share Knowledge Graph today. So Knowledge Graph is a new technology. All of your asset repository, your processes, service requests, integrations, it's already in ServiceNow. We want to leverage this knowledge. We are also, as Joe Davis showed yesterday in Bill's keynote, be able to connect to your SharePoint sites and many other third party sites, and we're gonna make that very easy this year. But if somebody's requesting a laptop, right? So say Amy is requesting a laptop, we already know that Amy has another laptop that she has not returned. Amy also has a phone that's active, some monitors, she's using some software, she's not using some software. We already know a lot about Amy from a technology perspective. In terms of integration, given they're all built in ServiceNow from an engagement layer like Jeannine talked about, we also know quite a bit that when somebody requests a 401k match, requests time off, booking travel, expense report in Concur, we know quite a bit. And then when it comes to people hierarchy, we also know the management hierarchy of Amy. Who does she collaborate with? Thank you for making me look handsome, Amy, I appreciate that. So we know all those details as well here. So this is what is knowledge graph for us. This is gonna fundamentally change that once you have knowledge, including in ServiceNow and third party systems, you apply that context to intelligence, and then you act using workflow. So Amy, I know that Jeannine talked about onboarding. Can we show what is our vision of onboarding done beautifully via intelligent workflows? - That's right. So we're really excited to share with you more about intelligent workflows. Our knowledge graph can deliver proactive, personalized, and most importantly, actionable AI experiences to employees. And those are intelligent workflows. Last year we brought you NowAssist. This year we're bringing you MyAssist. We are redefining the employee experience with MyAssist, which is the personalized assistant for work. So let's take Meera. Meera is a manager of an ITS team, and she's starting on her first day of work at ServiceNow. Now she opens up her ServiceNow mobile app to continue her onboarding journey. Now before she can get into the office on her first day, she needs her employee badge. MyAssist takes Meera's photo from her passport photo that she uploaded during her employment verification, sprinkles in some AI magic, and creates a beautiful badge photo. Meera happily uses this picture. It looks 10 times better than the original passport photo. MyAssist generates this beautiful digital badge, which gives her access to the office. But there's more. MyAssist also creates an avatar of Meera, which is a really fun employee perk. She loves it, and my assist publishes that image out to all of the applications that she'll use when she's at work. Next, Meera can also customize the look of her application, choosing from a company executive, to personify her AI assistant. After reviewing all these really fantastic looking executives, she chooses Chris Betty, because he is the leader of her organization. Chris is going to give Meera exceptional service. MyAssist proactively suggests that, based on her location at home, it's time to head into the office. And she is ready to go, she's got her digital badge. Okay, so let's fast forward, and Meera gets to the office before her orientation, opens up her new laptop, sees her beautiful employee center homepage with MyAssist ready to help. MyAssist has already given her access to the most important tools that she'll need to do her job. But Meera really likes to roll up her sleeves, and help out her team, and troubleshoot employee issues herself. She out that her team's using ScreenMeet, and MyAssist can just install it for her immediately. MyAssist suggests a few other things she can do to onboard, but there's been something that's giving Meera a little bit of anxiety on her first day. She already booked a vacation, she's not sure if she can take it. So using enterprise wide search, MyAssist answers her question, pulling together information from multiple sources, like knowledge base articles, the holiday calendar, even the files stored on box. So Maya's anxiety is relieved, she can take that time off and move back to her onboarding. She wants to know what's typical for team meetings, and MyAssist shares the typical meeting cadence for managers in her very same role, and completes the workflow by setting them up for her, plugging right into her Outlook calendar. But she realizes this is too soon. She's not quite ready to start her one-on-ones. She can go right back in the conversation, make an edit, and get the results she wants, with those meetings getting booked on the following week. Now one of the most important metrics for an ITSM team is mean time to resolution. And Meera is thrilled to see that MyAssist can generate this data visualization right there for her. This is so useful, she can pin it to her homepage, and that way she can see it every day when she starts her day. Now Meera also wants to know what her team is using to collaborate, and MyAssist can gather all the most important resources, and she can pin those too for future reference. Now one thing that MyAssist can do here is proactively detect that Meera does not yet have access to these documents. So MyAssist can access her proxy, going out and asking her team to grant her access to those documents so she can get into them when she needs them. Next in their conversation, MyAssist prompts Meera to send a welcome email to her team. It can easily draft a message. Meera can make a few changes, even adding a photo of her family for a personal touch, and that's a beautiful family. She can send the email, and then MyAassist knows it's just about time for her orientation. She better get up and start moving, and can show her exactly the way to get to that meeting. But Meera thinks of one last thing she wants to know. What should she know about her manager, Chris Betty? MyAssist summarizes a helpful snippet of information about Chris Betty, the chief digital information officer of ServiceNow. My Assist provides helpful links, like his LinkedIn profile, even his latest town hall presentation. And even better, MyAssist can pull together a number of photos from employee posts, seeing one that she loves in particular where Chris serves tacos to his team at lunch. What a fantastic culture he's built. Now that's worth favoriting. So that is the power of the Knowledge Graph that puts AI to work for every single employee. Thank you so much. Thank you very much, Amy Lokey. - That's wonderful. Personalization, Knowledge Graph in action. I have very special guest here today. I would like to welcome on stage founder, CEO, and president of Nvidia Corporation, Jensen Huang. - I want Service, Now. (CJ laughs) Thank you, Jensen. - I want. ServiceNow. - They agree with me. - Thank you, thank you. - We want our ServiceNow. - Yes, thank you. So Jensen, last year you were here, and it was still in the early innings, early stages of AI revolution. First thing is that we are delighted to be partnering with your teams. There are, as Joe Davis shared, multiple meetings every week. And we have collaborated, we are using you for building our models, training our models, and then inferences via A100 and H100s. Fantastic collaboration between DGX superpowers. H100 is six times faster than A100s. Little expensive, but six times faster. And then. - That's another way of saying six times cheaper. - Yeah. - It's okay. What's best for you. - Pretty awesome. Pretty awesome collaboration. And I know there's intelligent workflows. You have a high bar for your employees, you say you have to do extraordinary things to be extraordinary. How has ServiceNow rollout gone in the last one year at Nvidia? - Yeah, first of all, I'm delighted to be here. ServiceNow is the first enterprise software company that started working on a full stack generative AI platform. You know, we're going through a major platform shift, accelerated computing and generative AI. Every layer of computing is being transformed, from the computing infrastructure, the processors, to the acceleration libraries, to pre-trained models. NVIDIA's been pivoting into this direction for some time, and it was incredible that you saw the same vision, that we saw the same vision and worked together to do that. From pre-training the models, to aligning them, guard railing them, optimizing them for the performance, and the latency, and the cost, and the throughput, to the packaging of these things into microservices that we call NIM so that you can run it in every cloud, including on-prem. To building ServiceNow's only infrastructure, own infrastructure, so that you could deploy these, have a sandbox for your engineers and our engineers to collaborate in, as well as providing the service to your customers. I mean, across the entire stack you have really reinvented ServiceNow. It's really quite amazing to see. - And just so everybody knows, we also are built on ServiceNow. That's correct, thank you. Thank you, thank you. - Every morning I wake up, and the first thing I say is, "I want my service, now." And CJ just jumps right to it. - Yeah, thank you. So Jensen, you know, we have seen big shifts, personal computing, to cloud computing, to mobile, but the AI shift seems really, really accelerated and very fast. How should our customers think about where to use AI? Not just in general, you know, your healthcare companies, financial services company, manufacturing companies, life sciences, how should they think about AI and what will be possible with AI in next two to four years? - Yeah, one of the things that's really happening, CJ, is if you look at classically Moore's law advancing 10 times every five years, a hundred times every 10 years, over the last 10 years, AI has scaled a million times. First of all, you gotta take a step back and ask yourself, if something is moving a million times faster every 10 years, what should you do? And we're talking about the, for the very first time, the production, the manufacturing of digital intelligence. We used to manufacture electricity. The last industrial revolution was the manufacturing of software. Something that nobody understood that you could manufacture at scale. And we manufacture software at scale. And now we're manufacturing digital intelligence at scale. Yeah. - Now the rate at which this technology is moving is so, so fast. The first thing you should do is, instead of looking at the train from the side and trying to figure out why the train is moving faster and faster, get on the train. Because on the train it's not moving that fast. When you're on the train, it doesn't move that fast. Now, of course, one of the things that you've already shown in the work that you guys are doing, one of the reasons why the technology is being adopted so quickly is because we've transitioned from instruction driven computing, which is coding. Instruction driven computing, where very few people can do, to intention driven computing, which is intelligence, interacting with somebody through intention. And so even if you can't express, I mean, it's very often that when I'm trying to get help, I'm not able to explain exactly what the problem is and how to express the help. And so now with ServiceNow being able to have the intelligent assistant, you don't have to be exact in how you precisely describe the problem. ServiceNow understands what the problem is. And so I think that that's really something to consider, that every aspect of computing is gonna go from instruction driven to intention driven. Yeah, and you guys are done some amazing work. We wanna show you a very quick demo. As you know that Nvidia is really, really good, really, really good at graphics. And for a certain class of users, they would like voice to text rather than text. So Amy, quickly we can show the demo. - Okay, let's do it. Let's do it. So this is a customer service example, and we know that our customers wanna deliver seamless customer support, especially in the moments that matter. So imagine you're home and your wifi is moving incredibly slowly. It's down to a crawl, and you're the presenter in a critical meeting. You need service immediately. I need to contact customer support right now to see if I can upgrade. Now, normally I do this on my mobile phone. This is a live demo. So I'm gonna be talking to customer support through this headset right here. - Hi Amy, I'm your ServiceNow Assistant. How can I help you? - I'm having a hard time downloading files and my internet is sluggish, can you help? - I can see that your wifi service is working correctly. However, you have the 150 megabits per second package, the slowest connection speed we offer. Would you like to upgrade? - Yes, give me the fastest connection you have. - I can upgrade you to our highest speed of 800 megabits per second, which costs $60 per month for one year. You can save $60 a year, one month for free, if you choose this service for the next two years. - Sounds good, I'll take that two year package. - I have a credit card on file and can process the upgrade right away. Is it okay if I use the credit card ending in 6304? Yes. - I have upgraded your service immediately. Enjoy the fastest wifi we have. Thanks for choosing me to help you today. Have a good day. Thank you very much. Goodbye. That was a live demo between ServiceNow language models and NVIDIA's digital avatars. Jensen tells me the teams are working really hard for a certain class of users, for customer service, voice to text or text back to voice will be a real thing. I think obviously there's a lot of different ways to get and engage a digital intelligence. And of course text is one way. In the future, you'll be able to interact in a whole bunch of other ways. And one of the really interesting ways is digital humans. And so you're gonna see a lot of different applications of this in time. - Thank you, Jensen. Can you inspire us? With AI, five, 10 years out, understood the instructions, how faster things are moving, the cost is becoming cheaper, what would be possible with AI, and whether how work gets done or some of the big interventions that we thought are still not even possible? - Well, there's a whole bunch of things that the industry is working on that we're working on together. First, of course, you mentioned already multimodality, the ability to associate information that is embedded inside different modalities coming together so that you could find greater insight across the modalities. This is no different than cars that are using cameras, and radars, and lidars at the same time, trying to better perceive, better understand the environment through multimodality. That's one. Another is, today most of the AIs are one shot. You prompt it with something and it generates instantly their recommendation. In the future, AI will be multi-shot. It will be reasoning based systems. Just as we plan through various complicated scenarios, it will do some planning itself. And so you're gonna have fast thinking AIs, like we currently have. You're gonna have multi-step reasoning AIs that will be coming along. These types of AIs will become increasingly agentic. And you demonstrated some early examples of this as well. AIs that uses tools, AIs that works with other AIs, AIs that access information. They have privilege and access to, access control to, and so in the future, AIs are going to be like employees in our companies in the sense that our employees work with your employees. And so our AIs will work with other AIs, and we'll have consulting AIs, we'll have specialist AIs, we'll have generalist AIs. But one of the things that's for sure, ServiceNow is in such a great position because this is where our employees, the platform of our employees. This is the operating system of our company. And I've always hoped that Nvidia would become an AI. And I now realize that what I need to do is make you an AI and we'll use ServiceNow. That's true, that's very awesome. So Jensen, this AI revolution would've not happened if it was not for you and the teams at Nvidia, the greatest innovator. Thank you for being here, sharing your vision with our customers. Please thank Jensen Huang for being here today and inspiring all of us. - I want ServiceNow. I want ServiceNow. Ladies and gentlemen, that's the show. Thank you very much and we'll see you again. Bye. (upbeat music fades)