The Blueprint for Agentic Business | ServiceNow Knowledge 2026 Day 2 Keynote

Hello again and welcome back to the main stage. This is Knowledge 2026.

stage. This is Knowledge 2026.

Remember day one of knowledge? Feels

like it was only yesterday. That's cuz

it was. So,

let's keep the learnings coming.

[music] to lay out the Service Now blueprint.

for aentic business. Please welcome to the stage Amit Zavari.

[music] All right. Thank you, Adris. I always

All right. Thank you, Adris. I always

wanted that welcome. Welcome everyone.

How's everyone doing?

All right. How's everyone doing? Come

on.

Thank you for joining us at knowledge.

Hi Bill, what a great keynote yesterday and thank you for your vision and leadership.

Also want to thank our amazing marketing team and our CMO Colin for a great knowledge event [applause] and our incredible product and engineering team who have been working

very hard at the innovations you've been seeing this week. So, thank you for all the hard work you do and they've been very very busy as you can imagine. And

[applause] and Paul and go to market team, thank you for keeping us on our toes and keeping us humble and hungry. Thank you

for all the work you do.

[applause] What a year we've had. What happened in technology over the last year has exceeded everyone's expectation. Aentic

AI has gone from concept to a movement.

The pace of innovation has never been faster and the security landscape has fundamentally changed now that AI agents are working with us every day. But

there's one thing that hasn't changed.

Service Now is still about putting the power of technology in the hands of people doing the work. Before the world started talking about workflows, Service

Now saw the complexity buried in manual tasks and built a platform that connected people, systems, data end to end and to make the world work better

for people.

Now in the world of Agentic [music] AI, we are leading again by extending to every corner of your business with AI that delivers real results.

But we know the enterprise landscape is chaotic right now. That's because the average business has been running hundreds of apps, each with its own tech stack, data model, and security

policies. And the systems underneath

policies. And the systems underneath them were never designed to work together.

And that's what we call the patchwork enterprise.

Everyone is optimizing their own piece, but no one is orchestrating it as a whole. And that's why only 19% of

whole. And that's why only 19% of enterprise are actually getting value from AI.

But the companies that seeing real results, the ones with two and a half time x better outcomes from agentic AI are the ones running on a unified

platform with governance built in. And

only service now brings together data, AI, workflows and security on one platform. And that what makes us the AI

platform. And that what makes us the AI control tower for business reinvention.

And for over 20 years, as many of you know, Service Now has been the operational backbone of the enterprise.

And we run on our platform over 100 billion workflows and 7 trillion transactions every year, powering your IT, HR, CRM, security, and core business

processes and functions and for all the largest world's largest organizations out there. And at that scale gives our

out there. And at that scale gives our AI and agents productivity, governance, and trust from day one.

Today, we're giving you the blueprint for building an agentic business where AI senses, decides, acts securely at scale.

And since every blueprint starts with the people who build, that's where we'll begin with the engineers, developers, and creators who turn the architecture

into something the business can actually use.

As you have seen nowadays, innovation has never moved faster at the edge.

There are a lot of tools out there like VS Code, Codeex, Claude Code, Figma and others where you can build products and prototypes very quickly. But because of

this prototypes and prompts and agents are scattered across teams and tools and the result because of this is dangerous gap between prototypes and production

and today we are closing that gap by building anywhere and running on service now. You go from prototypes to

service now. You go from prototypes to measurable business value.

And here's how we're making that real.

If you're building in Service Now studio, build agent is right there with you. Purposemade for Service Now

you. Purposemade for Service Now development and the most powerful building experience on our platform. And

if you live in other live in other tools, no problem. Build agent meets you there too with core skills now available through the service now SDK.

But some of you might be thinking if you have the engineering talent, you have access to foundational models using APIs, why build on Service Now AI platform

when you can assemble those pieces yourself. And here's why. With Service

yourself. And here's why. With Service

Now, you can use any environment or coding agent. And you'll get the full

coding agent. And you'll get the full power of the platform with enterprise ready governance, security, and compliance at your fingertips.

And as as we announced yesterday, Service Now action fabric extends the same platform power to every AI tool already running across your enterprise.

The build agent skills and MCP integrations are already live today.

And to learn more, make sure you attend the developer keynote at Creator Con tomorrow. It will be very deep dive into

tomorrow. It will be very deep dive into the technology we are delivering for you today.

So now please welcome Danielle to show us what build agent would look like inside an enterprise environment for with CVS as well.

Thank you Amit. Thank you.

All right. AI has fundamentally changed how we build applications. Let's take

vibe coding as an example. It's about

using your natural language and your intuition to generate code. But let's go further than just code generation. Let's

have a look at how we can build full applications with the right workflows, data and governed logic. Now I am a developer from CVS and the business have

just approached me and they would like to add a new benefit for pet insurance to their benefits application. As a

developer, I want to work in the coding tool of my choice. And in this scenario, I'm in Clawude Code. So, I'm going to quickly enable the plugin for Fluent at

Service Now and use a very natural language prompt to add pet insurance to the benefits application. Let's get this off and running. And this right here is

Fluent in action. This is Service Now's modern coding framework where we have opensourced our build skills and platform knowledge so that you can vibe

code a service now application in your coding agent of choice. But what you will notice is that this application is built with the right data model from the very beginning. Right components and

very beginning. Right components and data model. This application has even

data model. This application has even been ported into a Service Now sandbox environment, isolated away from production and ready for a governed

deployment process. So I think on that

deployment process. So I think on that note, let's jump into Service Now Studio. And in the corner, we can see

Studio. And in the corner, we can see our extra care benefits application was just updated and I'm getting a number of AI recommendations for how we can

improve this application.

Now step one was to build the application but step two is to make it autonomous. So the platform is

autonomous. So the platform is recommending that we build an enrollment AI agent. It would understand an

AI agent. It would understand an employees eligibility, manage the endto-end enrollment process without any human intervention. I think it sounds

human intervention. I think it sounds like a great idea. So let's go ahead and use build agent to create this custom AI agent for our application. This again is

taking the app from just responding into a mode of acting autonomously just in time. Now our AI agent is ready. It has

time. Now our AI agent is ready. It has

the right instructions, skills, and roles to act autonomously. However, we

aren't done quite yet because we need to test it and make sure that this is actually production ready. So I've run a few automated test scripts and we have a

readiness score of around 90%. So, a

couple of changes which will move into the next release, but I think we're we're on track to submit the deployment request. So, as we review our release

request. So, as we review our release notes, as I like to say, let's get this baby up to prod and deploy this application so that our CVS employees

can benefit from the pet insurance benefit. Now, I I have a dog at home, so

benefit. Now, I I have a dog at home, so I want to I want to learn more about this pet insurance benefit. Let's see it see it in action. It looks like the AI agent for eligibility is checking

whether I'm eligible. And I am.

Fantastic. So, my little Chihuahua Frankie, uh, let's get him added to the to the uh, pet insurance right there.

There you have it, everyone. We have

changed the game for app development.

And at CVS, their apps don't wait. They

sense, decide, act, and they are governed on the Service Now platform.

Amit, on that note, back over to you.

[music] Thank you, Danielle. It was incredible to see how Service Now auto, build agent, and the coding tools come together to ship enterprisegrade apps

faster than ever. And what Danielle showed to you today is one example of what's possible. At any given moment,

what's possible. At any given moment, the Service Now platform is orchestrating work across the world's largest enterprises, the systems they run on, and the teams that depend on

them as well. It's powered by four interconnected capabilities that turn AI inside workflows into autonomous

execution. One platform to sense what's

execution. One platform to sense what's happening across every system and every workflow. Decide with the full context

workflow. Decide with the full context of how your business actually runs.

act in terms of through the workflow they execute across departments, systems of record and cloud and secure every agent, every identity and every asset.

That's your blueprint for agentic business. Now let me hand it over to

business. Now let me hand it over to Gorav to take you into how AI that senses.

[music] [music] Thank you.

So at Service Now, at Service Now, we believe that for work to be truly autonomous, you need nothing short of a complete sensory nervous system. And

this sensory system must connect all your data wherever it lives, control it with enterprisegrade governance, contextualize it with enterprisewide

intelligence, and then converge that contextualized intelligence right into the flow of work.

So first connect workflow data fabric is fundamentally architected for the agentic era. connecting all your data

agentic era. connecting all your data wherever it lives and allowing humans and AI to treat a vast patchwork of

enterprise data as if it were one.

See, most data fabrics are primarily built for decision support, for insights. Workflow data fabric is built

insights. Workflow data fabric is built for insight and action with read and write capabilities and support for all

types of data, structured or unstructured, internal or external, static or streaming, still or sparkling.

And that's another crucial distinction.

We embrace the system of record and data platform choices you've already made.

You can but you don't have to move the data into service. Now look I know there are others who are playing for data

gravity but for us what really really matters is knowledge gravity. We even

have a conference with that name.

Workflow Data Fabric already offers 250 plus connectors and we're expanding our reach even further with 100 plus new zerocopy connectors so customers can

access data wherever it resides. No

replication required. With full support for MCP clients, so our AI agents can work with any MCP enabled source. With

auto for workflow data fabric, so customers can describe what they want in plain English and let AI build the integration for them.

But look, connected data is not the same as AI ready data. To ensure AI decision accuracy, you need data that's fully visible and governed throughout its

entire life cycle. You need tight control of data. And this needs to be non-negotiable.

And so that's why we're introducing Service Now data catalog which delivers native metadata management, data

lineage, privacy, and ultimately trust so that humans and AI agents can now instantly discover and use curated data products safe in the knowledge that

these data products have your enterprises seal of approval.

That's great. But getting data AI ready isn't a one-time activity. Enterprises

have to keep it AI ready. And so to ensure ongoing AI readiness of data, we're taking our data control capabilities even further with a complete autonomous data governance

capability.

And so we'll be delivering data quality, observability enrichment policy management, all unified inside the Service Now AI platform. And we'll offer this through a combination of Service

Now and partner products in our massive workflow data network.

On to contextualize.

We've connected the data. We've put it under control. But to really drive

under control. But to really drive trusted decisions and actions, AI agents must have clear line of sight to business context. Public LLMs are blind

business context. Public LLMs are blind to this, but Service Now isn't. With a

100red billion workflows a year running on our platform, no one is better able to understand our customers business context than us. And so we bottled that

magic up into context engine. So our

platform of platforms, as Bill likes to call it, now has a graph of graphs. And

that graph brings together our knowledge action access asset and decision graphs all anchored on our powerful CMDB.

And right into that context engine is built our marketleading semantic layer that we have used as a foundation for a new solution that we are announcing

called autonomous data analytics. This

is big. So think conversational and natural language analytics to guide the what happened, what will happen and what should I do types of decisions that AI

and humans have to make every day. And

so it's fully autonomous. You know,

think an embedded always on AI analyst working tirelessly on your behalf, surfacing insights, interpreting enterprisewide data in context, spotting

issues, detecting outliers, providing recommendations, even taking action.

And finally, converge.

Today, enterprises analyze in one system, act in another. We decided to unify both at the database level with Raptor DB. We have Raptor DB standard

Raptor DB. We have Raptor DB standard which is freely available to everyone and a premium Raptor DB Pro that delivers even greater performance and

scalability. And today I'm excited to

scalability. And today I'm excited to share two important new capabilities we're adding to Raptor DB Pro.

Live Archive, cost-effective archival solution for Service Now that allows you to then seamlessly query across hot and cold data. and live connect which lets

cold data. and live connect which lets you point your existing BI tool against Raptor DB Pro for real-time analytics with no ETL or data movement.

So there you have it, the four C's of our sensory system. And if you'd like to learn more about them, please join me at the solution stage A at 1 p.m. today.

And now to tell us more about the decide component of the AI control tower, it's over to you Nad.

[music] [music] Sense is something my wife says I don't have enough of, but gives AI live access to data across the entire enterprise.

But seeing and understanding data are two very different things. The large

language models powering AI today are extraordinary at understanding language and reasoning through complex problems. They were trained on the internet. So

they're incredibly good at finding patterns in text, code, knowledge, and the fundamental substrate of the internet, cat videos.

But what they're missing is the context of how your specific business actually operates, leading to the unfulfilled promise of autonomy.

For example, when a server goes down at a financial services company and my one desire is to have it restored,

most AI models will believe when I say that I want it that way.

[cheering] Context engine will tell me why I want it that way. The heartache when 14 downstream

way. The heartache when 14 downstream services broke the last time we did it that way. And whether the risk team is

that way. And whether the risk team is going to tell me it ain't nothing but a mistake to do it that way again.

[applause] We told you Backstreet is back.

That's because context is different from data. It's the history of every decision

data. It's the history of every decision your business has made. How that

decision got made and what happened next. The more decisions, the more

next. The more decisions, the more history, the more context.

So the real differentiator is whether AI has the business context to make decisions that you would trust it to make on your behalf in the middle of a

critical workflow.

This is where Service Now's 20 years of modeling enterprise operations becomes a compound advantage.

We've been mapping the extensive relationships between people and processes and assets and services and policies across every touch point

building the enterprise intelligence that AI now gets to operate on. This

includes our knowledge graph. It

includes the CMDB relationships, the cyber asset graph and the decision graph.

Context engine unifies all of that into a single organizational layer. It takes

a snapshot of every factor involved in a decision, including what was happening, what action was taken, and what outcome resulted.

Once AI has full business context, inference becomes proactive.

And because every decision, every action, and every outcome feeds back into the system, it gets smarter over time.

That's the difference between a model that can reason generally and a system that reasons about your specific business and improves every time it acts.

Now that we know what context engine does, I'm going to turn it over to the one person who can make context engine larger than life. Please give it up for

Shardi Patel.

Thank you, Nshad. We will see you back on stage tomorrow at Backstreet Boys.

All right. Now, we all know that AI is only as smart as the context that it has. And most AI is pretty good at

has. And most AI is pretty good at suggesting, but with context engine, it goes beyond suggesting to taking action that's both accurate and precise. Let me

show you what I mean. I'm going to step into the role of an HR business partner.

and someone on my team that I support, Sarah Johnson, wants to make a lateral move in the company. All I'm going to do is type one sentence here into auto.

I need to process a role transfer for Sarah Johnson from finance to business operations effective Monday at 10 a.m.

I'm going to step away and let Context Engine handle the rest. Now the first thing that's going to happen is user graph is going to pull together a complete picture of Sarah starting with

her role, her tenure and her location, but also more importantly systems that she's supposed to access. Now four years in finance for Sarah means four years of accumulated access to things like

workday and ADP payroll and her finance SharePoint. User graph also knows what

SharePoint. User graph also knows what she's going to need on the other side.

SAP Teams and her new business operation SharePoint. So, user graph has basically

SharePoint. So, user graph has basically mapped who Sarah is today and who she needs to be on Monday at 10 a.m. and fed

all of that over into context engine.

Next, knowledge graph is going to surface the right rules and policies to keep this transfer moving forward.

It's showing me that access to Sarah's finance systems has to get turned off first because that is company policy.

And every removal needs to be documented and logged for socks compliance. And for

every system on Sarah's list, both old and new, knowledge graph has surfaced the right playbooks for provisioning and deprovisioning access.

Okay, so we just saw with user graph that that's showing us the systems that Sarah's supposed to access. But with our security graph, this is going to show us the systems that Sarah can actually

reach. And as we all know, those two

reach. And as we all know, those two things are not always the same. So,

Security Graph is mapping every permission path for every connected system that Sarah has. And yikes, her blast radius is pretty significant because she still unintentionally has

access to things like executive compensation data and board reporting files. But content context engine has

files. But content context engine has already mapped all of it for removal and her access is cleanly removed and fully documented.

All right, so we're on to our fourth and final graph here, which is decision graph. And this is where the decades of

graph. And this is where the decades of work running through the Service Now platform are really going to pay off. It

is taking all of that institutional history and turning it into intelligence that's calibrated to how your business operates. So, what do I mean here?

operates. So, what do I mean here?

Decision graph has analyzed 23 prior finance to business operations transfers in my organization, and it's picking up on something that most managers wouldn't

even think to check. It's showing me here that employees who complete two specific courses, business operations fundamentals and SAP reporting

essentials, onboard and ramp 50% faster than those who don't. So, of course, I'm going to enroll Sarah today.

I want to pause here for a second because this is so powerful. Decision

graph isn't just picking up on prior patterns. It is proactively solving

patterns. It is proactively solving problems for me that I didn't even know I had.

All right, let's fast forward. It's

Monday at 10:00 a.m. and Sarah is walking into her new role and everything just works. Her old access is gone. her

just works. Her old access is gone. her

new tools are provisioned and she is well on her way to becoming a smooth business operator.

Now, I know we just went over quite a few graphs, four in fact, but no single one of these graphs was going to give us enough information to make this transfer

happen. It's only when we can reason

happen. It's only when we can reason across all of that simultaneously, your people, your policies, your CMDB, your prior decisionmaking history. That is

when we get intelligence that can actually run your business and that's context engine. And nobody is going to

context engine. And nobody is going to give your AI more business context to work with than service now. Now to share how our platform becomes even more

powerful with our partners. I'd like to welcome back Ahmed on stage.

[music] Let's give it up for Gorav.

Let's give it up for Gorav Nshad and Shardi. Great job. [applause]

Shardi. Great job. [applause]

So we have shown you how the platform senses your data and decides with full business context. The next natural

business context. The next natural question is how do the best models in the world plug into that? One of the ways we make this possible is through strategic partnerships and an open

ecosystem.

Any AI model, any data source, any cloud, any hyperscaler and any tool as you saw earlier. One of those partners is Google cloud. So let's welcome Karthik Naran to join me on stage to

talk about what we're building together.

Car take [music] [music] thanks for joining us today.

Thank you for having me here.

Of course, last week uh at Google Cloud Next, you guys unveiled a lot of lot of new capabilities and we also unveiled a significant expansion of our partnership in terms of how we bring autonomous

operations to the enterprise. Can you

just share about how are we moving together and the shared vision we have into reality?

Yeah, first of all, we are so delighted to be uh partnering with Service Now and I've always felt that the the level of synergy that um our two platforms bring

together to make things frictionless for our customers is is uh unparalleled. I

think the market has moved um from the stage of asking you know whether should we use AI and how do we use AI to you know adopting AI at scale and scale is

generally misconstrued uh where people sometimes think it's about volume it's not about volume it's about adapting AI not just as a feature in an existing

workflow but to reimagine workflows with AI you know as a core architecture to it And that's where I think our partnership is bringing that you know agentic

enterprise vision to life and we are doing that in three ways with you.

Number one is uh everybody's talking about uh Google's full stack AI capability and we are bringing that full stack capability to the service now

ecosystem and that is from our hyper uh computer infrastructure to our models to our agent platforms to uh the agentic

data cloud all of these capabilities are plugged in and it's the connection is not just a company to company one you know pipe conction ing we are connecting

it at every level of the product stack um so that um you will be able to seamlessly integrate our models into your workflows. uh the data platforms

your workflows. uh the data platforms talk uh seamlessly not you know while there are agentto agent protocols those are thin pipes but if you really want to

run a workflow seamlessly almost like a first party experience and that's the the core when when we talked and you know Bill uh and I talked we always talk

about how do we give a first party experience to our customers and that is the intention that we are doing at that level the second thing is the unified governance There's a seamless integration between

the um Gemini enterprise agent platform and your AI control tower. What do I mean by it? You when somebody creates an agent and registers that agent, that

registered agent automatically gets registered in Gemini Enterprise and it shows up in the console of Gemini Enterprise. The second thing is on

Enterprise. The second thing is on Gemini enterprise somebody can directly create a service now agent and it's registered on both places so that there

is seamless governance that organizations can have especially the CIO CTO co uh that's sitting here have

agent sprawl as a challenge and the third is uh we are breaking the data barrier so we are directly connecting our agentic data cloud with um you know

your data platform so that there is seamless integration for thick context.

Everybody is worried about long context.

There is also a concept of thick context. You need to be able to process

context. You need to be able to process so much data real time uh to be able to give the experience. So that's how we are doing that.

Thanks for sharing that. I think this whole first party experience where agents can be discovered across the all the different technologies out there and bringing that together I think has been a great partnership as well. This leads

directly into my into the two major innovations we are also now introducing to the market this week. right around

customers and the field operations around this idea of intelligent gateway vision for CX and the native integration of Geminina live into service now field service management and that has been a lot of co- innovation between our teams

just get great to get your idea in terms of how you taking this to market as well as what the what the customers and technicians on ground should expect out of that yeah I think the fundamental challenge

that AI is solving is something called as context blindness that gets created when either a customer is interacting with an enterprise or the field

technician is working on a problem. uh

and the the reason for it and let's take the example whether somebody's calling a a support and you know they are switching channels from chat to voice to

video or they had a conversation and they are coming back after 1 hour you know there is a lot of context and all the emotion that was discussed during

that call or that interface is lost and the reason for it is all traditional systems record the end state of a transaction. It records what somebody

transaction. It records what somebody ended up buying or what was the end resolution of a product. It does not track the entire contest or the

traversal that takes to the end point.

AI's ability is to track and record the entire traversal uh so that you understand the emotion so that you don't feel like you are repeating. Um that is

what you know we are trying to solve it through these two innovations that we're creating. The first one is uh for

creating. The first one is uh for customer experience you know we have a product called Gemini enterprise for customer experience. What we are now

customer experience. What we are now doing is that is going to create and capture and hold the context across all these channels. Um but it is going to be

these channels. Um but it is going to be directly integrated with you know your knowledge graphs so that when somebody

is talking say for example um you are working with your you know calling your telco and you're saying while you're traveling abroad from the airport you're asking for certain changes more data

certain changes in your plan and so on so forth and now you have to board and you cut the call and you log in to your app after the flight takes off. Usually

this context is lost right um and now what we are doing is that entire context is maintained and it is mapped with the knowledge graph on service now so that

the actual information is pulled from service now and the action is taken on a button click so that's the integration from a product level Gemini enterprise for customer experience is going to be

directly integrated the the um the knowledge engine for that is going to be service now for a lot of aspects for customers where the data is already

there uh in service now and we also feel there's a lot more use cases where that information should actually be in service now. So we are going to push

service now. So we are going to push that uh from that perspective on uh field service. This is a a huge problem

field service. This is a a huge problem because it's the perfect example of a lose-lose proposition where uh an

inoperable equipment is the the biggest you know um dissatisfaction for a customer and a truck roll is the most

expensive thing for an enterprise and as a result first touch resolution is the most important metric that everybody needs to have. But you know all the time

when an a technician goes in there's almost like a physician kind of a scenario the most experienced person is able to do the resolution but anybody

other than that results in multiple um times you know the the field technician needs to visit. Now what we are doing is we are integrating Gemini live reasoning

with you know audiovisisual context with service now which means that somebody could you know make AI watch the equipment performing. It can hear some

equipment performing. It can hear some sound you know a noise getting created before a break and there could be a smell that is coming from some burning or something that the technician could

talk and activate this. What does that do? Step two that activates the sub

do? Step two that activates the sub agents in um service now with skills exposed for whether that hitchback system or some other equipment and

actually provides the wiring diagrams or cited repair you know solutions from the experts so that you could solve the problem and step number three is a lot

of time is wasted on you know after service records management so all of that could be done by Gemini into service now directly and also in certain

cases using nano banana you can create new uh diagrams and manuals and so on and so forth that could be created so this would create a seamless experience

but it solves a problem that's that nobody wants to have at the field thank you Karthik for getting into the details of how we integrating a lot of our products together and that's the power of the great partnership which is coming

together to turn AI into valuable execution and making it really a first party experience so thank Thank you for that partnership. To take us further

that partnership. To take us further into AI that acts, please welcome Kelly to the stage. Thank you, Kelly. [music]

Thank you.

[music] So, we've talked about how the platform sensus data and how context engine helps AI decide with full business

understanding. Now we get to my favorite

understanding. Now we get to my favorite part. What happens when AI actually does

part. What happens when AI actually does the work? Taking action has always been

the work? Taking action has always been at the core of Service. Now we are the system of action. And now we've extended

that DNA to the agentic business with AI that executes. We've gone autonomous.

that executes. We've gone autonomous.

That means every capability I'm about [snorts] to show you runs inside the policies, the permissions, and the governance your organization already has

in place. So you get full visibility,

in place. So you get full visibility, full audit trail, and full control.

That's winning all day long. So let's

begin where we started with a solution that still runs the enterprise today. IT

service management. This is where most of you began your journey with us and we thank you for that. It's where CMDB was born. It's where workflows first

born. It's where workflows first connected with real business services and where our platform earned its

reputation as the system of action. AI

agents use CMDB in our knowledge base to diagnose issues, determine root cause, and handle IT requests from first touch

to resolution autonomously before the business even knows they're an issue. And as a CIO, I am so here for

an issue. And as a CIO, I am so here for that. I live it every single day. At

that. I live it every single day. At

Service Now, 90% of our IT support requests are handled autonomously. That's what it means to go

autonomously. That's what it means to go from roadmap to reality. And yes, it is live in production today. Thank you

team. Love. Yes, team work.

[applause] And to learn more about autonomous IT, head to the spotlight session right after this keynote. The team will be there to take care of you. And what's

even more exciting is what works in IT, it works in every other corner of your business, too. HR and employee

business, too. HR and employee experience. AI agents remove the

experience. AI agents remove the friction from onboarding workforce operations and everyday employee

requests. in security and risk. AI

requests. in security and risk. AI

agents detect threats, close compliance gaps, and remediate vulnerabilities real time inapp development. Anyone can build

on the platform faster with AI code generation, accelerating workflow creation, and testing. Amazing.

And you've heard it yesterday. We've

introduced autonomous CRM. Now a sales rep can see the full picture of a customer in one single view. A support

case arrives pre-triaged with contacts for the human support agent and orders flow from quote to

fulfillment without manual handoffs. So

what you get are AI agents working on the same platform with the same data in these same governments. That's huge. AM

I RIGHT?

RIGHT.

And all of this comes through something I'm really excited about. It's

autonomous workforce. These AI

specialists are agents trained to do a specific job. They have defined roles.

specific job. They have defined roles.

They're assigned to existing teams. And they execute workflows end to end. And

the best part, they're not starting from scratch. Autonomous workforce is

scratch. Autonomous workforce is possible because of everything we've shown you over the last two days, combined with all of the years of

institutional knowledge already built into the platform. So, what you're activating is what you've already invested in.

And we announced yesterday we've got 20 new AI specialists to cover every corner of your business. So IT professionals

out there, I got you. HR, we got you.

Customer service, you get one, too.

Finance, of course, there's one for you.

Security, you know it, we got you. And

many, many more. They all operate within your existing assignment groups and your existing governance. That's the beauty

existing governance. That's the beauty of the platform. You truly get AI that thinks and workflows that act.

Now, let's focus on our awesome customers. CVS Health, who mentioned

customers. CVS Health, who mentioned earlier, has delivered over 2.5 million AI conversations, saving

valuable time for patientf facing and customerf facing work. Here's what I love about it. Everybody wins. Now, we

want to show you what this could look like at CVS Health. Let's go, Nikki.

It's time to show the amazing demo. COME

ON UP.

ALL RIGHT.

Thank you, Kelly. Thank you for that amazing setup. My name is Nikki Patel,

amazing setup. My name is Nikki Patel, and I'm joined on stage by the lovely Miss Britney, who's going to help me drive the demo. Now, as the product lead who led the team to build the first

autonomous worker, I am thrilled to be here on stage to show you how this is going to transform and elevate your teams. Now, let's go ahead and explore what that means for you and your teams.

For the purposes of this demo, I'm an IT service desk manager at CVS Health. Now

I start my day by going to my homepage and I can already see the L1 IT service desk AI specialist I onboarded yesterday in action and for the first time. The

tedious work is getting done even before I realized it exists. Common tickets

that used to sit there for days if not hours all resolved. Though I'm noticing something else. Software access

something else. Software access incidents are spiking fast. They're

piling in. The cues are piling in. and

all the capacity I had left over is getting quickly absorbed. Now, this is usually the moment my team starts to lose control and fall behind. But this

time, it doesn't. The AI adviser steps in. The AI adviser proposes we add the

in. The AI adviser proposes we add the software asset management AI specialist to the team. Purpose-built to absorb this type of demand. Now, it's

continuously scanning my instance, my environment, and tells me exactly where to add the next AI specialist. I didn't

have to go looking for capacity. It

showed up right when I needed it. Now,

as you can see, the AI adviser explains the role of the software asset management, Sam, the AI specialist, and why it matters to me based on the context and data it's mind. Now, when

access requests spike, the bottleneck is familiar. There's no clear view of

familiar. There's no clear view of available licenses. And let's not forget

available licenses. And let's not forget about the arguous process of manual fulfillment. But the software asset

fulfillment. But the software asset management AI specialist removes that friction. Though, what if I don't want

friction. Though, what if I don't want it to handle all of it? What if I want to be in control? And in the words of Nshad, what if I want it that way? So,

every AI specialist is fully configurable. I define what it can and

configurable. I define what it can and can't do. And these configurations

can't do. And these configurations directly shape its behavior. No prompt

rewriting, no instruction tuning required. I craft the experience I want.

required. I craft the experience I want.

And the AI specialist adapts. So, every

interaction feels natural and consistent with the way my team operates. Now, how

does this AI specialist actually do its job and how is this different than a single AI agent? I know we've been throwing a lot of terms at y'all during this entire keynote. So, let's go ahead and take a moment to define what this

means. Now, an AI agent executes

means. Now, an AI agent executes specific actions. This is where you want

specific actions. This is where you want human on the loop in the loop and it requires a lot of creative judgment. You

only want AI agents to be doing specific actions. Now, an AI specialist delivers

actions. Now, an AI specialist delivers an endto-end outcome. It's an

coordinated team of AI agents, each trained for a specific job, working together as one team member. And they're

not acting in silos. They're

collectively running within your workflows, your policies, and your rules. Now, imagine having to go build

rules. Now, imagine having to go build all that yourself. As you can see, anyone can spin up an AI agent these days. But what's hard is making it

days. But what's hard is making it reliable, governed, and consistent every time it acts on your business. a system

of agents coordinating decisions, passing context, enforcing policy, and staying in sync as everything around you and your business keeps evolving and

changing. The complexity doesn't stay

changing. The complexity doesn't stay flat, it compounds. But Service Now has already solved for me what's hardest to get right. They've built, trained,

get right. They've built, trained, tested, and are governing this AI specialist for the job out of the box.

designed to operate in my instance on my data and my policies and just like that it's onboarded. So now we have two AI

it's onboarded. So now we have two AI specialists working together. We have

the L1 IT AI specialist resolving common L1 incidents and then the software asset management AI specialist resolving software access and licensing. So let's

go ahead and see what that looks like in action. The moment the incident arrives,

action. The moment the incident arrives, the L1 IT AI specialist starts triaging it, but it quickly realizes this isn't an incident. This is an actual software

an incident. This is an actual software access request. So what it does is it

access request. So what it does is it creates the request and routes it directly to Sam the AI specialist. Now

Sam the AI specialist quickly realizes there's no licenses available. Zero

capacity in licenses and 52 employees are blocked. And at CVS, every minute

are blocked. And at CVS, every minute blocked is a minute they're not serving their employees, partners, and customers nationwide. So Sam the AI specialist

nationwide. So Sam the AI specialist acts by first reclaiming the idle licenses and then procuring the delta and in parallel sending approvals to finance in procurement and proactively

updating all 52 employees. Now the

moment approvals is cleared access is provisioned. Now what would have taken

provisioned. Now what would have taken weeks is resolved in a matter of minutes. This is what a coordinated

minutes. This is what a coordinated autonomous workforce looks like in production. Work doesn't wait in cues

production. Work doesn't wait in cues anymore. It gets done even before anyone

anymore. It gets done even before anyone notices it's there. End to end across your workflows within your guardrails.

Now, next up, let's hear from one of our amazing customers who's helping people on a path to better health. Take it

away.

[music] Our purpose is to simplify healthcare.

[music] One person, one family, one community at a time.

We have [music] an incredible responsibility. CVS Health serves 185

responsibility. CVS Health serves 185 million customers, members, and patients and 300,000 colleagues caring [music] for them.

That's right.

Deliver better service for our customers. We have to deliver better

customers. We have to deliver better service to our colleagues. This is the story of CVS Health and the Service Now AI platform at work and at scale.

We [music] often talk about human connection and the technology we build is with that in mind. With AI, [music] the industry changed and Service Now evolved with us. I remember when we first got Service Now and [music] Boy

works at CVS Health. We call it Arty.

Nice to meet you, Arty.

Before I spent a lot of my time looking up information, but now whatever I'm working on, whatever I need to get done, I know that answer is right here. That's

where [music] Service Now employee works became more of a thought partner than just a tool. What we love about Service Now employee works platform is that [music] you can ask it, it'll tell you it, and then it'll go do it. You're all

on the same user experience.

It takes action.

Yeah. So, say I'm having problems with [music] my computer, right? With Service

Now, employee works, it can go in and diagnose that answer, fix it, or put in a [music] ticket for me inside other applications.

That's right. So, I'm able to stay in the flow more to focus on more strategic [music] work. So, for me, that

[music] work. So, for me, that transformed my work. I work smarter, faster, [music] better. If you scale that to 15, 20, 50 colleagues, all of a sudden, you've recovered 5,000 hours

[music] of productivity. That's real

time. That's real money.

That's big.

We've had over 2 million conversations with the product in just about a year.

And our return rate is over 75%. [music]

That means that folks are coming back to the platform over and over again.

Wo. No matter who you are as an employee, there's work that you have to do and there's work you want to do. So

imagine in a store a customer comes in and says, "Excuse me, my child is at home with a flu or or whatever it is."

Where people have time to listen, [music] to guide them, to help them.

Thanks.

That's the moment.

Bingo.

That's [music] why we're here.

When CVS Health works, the world works.

[applause] So we have looked at what's possible with CVSL across multiple demos today.

They created a single employee experience with Service Now employee works. One front door for everything

works. One front door for everything across IT, HR, procurement, corporate communications and store operations.

Work moves faster, stores runs more efficiently, and teams can focus on what CVS Health is really about, delivering care. With that, I'm excited to welcome

care. With that, I'm excited to welcome Alan Rosa from CVS Health to join us on stage.

[music] [music] Hey, Alan. Welcome.

Hey, Alan. Welcome.

[music] I love the shoes.

Thank you. My wife picked them out.

Always listen to her. [laughter]

So Alan in just 9 months uh your team has unified your whole employee experience with a single AI platform.

Walk us through what made that possible and the move that that fast at scale and the value you're seeing so far.

Well look every story starts with a bit of gratitude and um you go along with people. I'm fortunate to work with the

people. I'm fortunate to work with the best colleagues best engineers in the world and that's our CVS health engineers and colleagues that make this

possible. So my gratitude goes to them

possible. So my gratitude goes to them [applause] you know and to extend upon that you don't you don't do it alone our partners our account team from service now who

are with us every step of the way our partners from deote KPMG RSC consulting and tech systems are in the audience today supporting us cheering us on so

our gratitude goes to you as well [applause] but let's Let's get down to the tactics.

Um denominator here is discipline and focus.

You got to stay on task. The numerator

is defining your problem. Making sure

you get rid of the tech debt. You're

going clean, not fast. Y

and finally, your architecture matters.

Let's break that down a little bit, right? We define the problem.

right? We define the problem.

Frictionless and disintermediation.

We want to make sure that our colleagues know that problems are going to get solved. They're not going to use the

solved. They're not going to use the tool. the tool doesn't address the

tool. the tool doesn't address the issues that they're having. So we

settled on what the problem definition was and then we aligned internally. This

was not a product selection. It was a change in operating model, the way people worked. And then we aligned

people worked. And then we aligned ourselves internally. Technology, HR,

ourselves internally. Technology, HR, finance, procurement, store operations.

There's no sense in building a front door if people aren't going to walk through it.

Right?

And then secondly is let's not go fast, let's go right. Let's remove the tech debt. So we initiated a seven-month

debt. So we initiated a seven-month project called Greenfield and we essentially rebooted the platform all of it and we were able to do that in such a way without disrupting the business but

reenabling it so that when we launch these new capabilities Service Now is ready for it and we did that together.

Then lastly is the architecture enabling the conversation. Move works handles the

the conversation. Move works handles the intent of the conversation, but Service Now does all the heavy lifting, the approvals, the workflow, the system of record. Now we're moving from a

of record. Now we're moving from a conversation to an outcome. So what do the results tell us? 220,000

people using the platform, 4.65 million plugins that are being leveraged. We're

talking about removing 255,000 calls to our service center operations.

Those are real outcomes as we say in CBS Health. Boom.

Health. Boom.

Wow. [applause]

It's a very very impressive journey and I think to be able to do this in short amount of time and to be able to get this volume of usage is very impressive.

The grateful thank you for the partnership and the work we do together as CESU as well. I mean you're doing impactful work of securing AI that touches the care of 185 million Americans. So what do you think it takes

Americans. So what do you think it takes to kind of do that kind of trust at enterprise scale? So, this is a slightly

enterprise scale? So, this is a slightly different conversation. Um, trust is the

different conversation. Um, trust is the only capital that CVS Health really has.

We lose that, we stop functioning as a business. You trust us for a reason and

business. You trust us for a reason and we take that very seriously. So, we

engage in an AI conversation, we're already going to do things that make sense that we can deliver that value.

Everything we look at from an AI perspective is about privacy, legal, security, and governance or we're not doing it. Full stop.

doing it. Full stop.

Mhm. Right. So we undertake this, we think about okay um can we put the guardrails in place to ensure that we can protect you, our members, our

patients and our customers. Now as a seesaw, the other thing I think about is that AI is breaking every single mental model we have when it comes to security.

You're thinking about prompt injections, you're talking about data leakage, you talk about data mod uh validation, uh model validation, and it's a perpetual evolution. One of the reasons that we're

evolution. One of the reasons that we're partnering with Service Now, and we're excited about control tower because we're building a warrant system that's not as human dependent. That's too slow.

We need to innovate at horizontal scale, especially for a company of our size.

This is the optionality you're now giving us, and we can build this together. And lastly, zero trust.

together. And lastly, zero trust.

Mhm. Trust nothing, verify everything.

We have to make sure that we're taking the right steps. The last thing, and it's the most important part, engagement of your board, your ELT, and your stakeholders. Make sure you're bringing

stakeholders. Make sure you're bringing them along for the ride. The frameworks

are there. We all know NIST. We We all want to get out the ISO frameworks.

Listen, we all want to get up the MITER stack, right? You're doing the right

stack, right? You're doing the right things, right? Tabletop exercises,

things, right? Tabletop exercises, penetration test that gives you the remediation road map. follow it right but understand where your duty care come

from right the MON standard inry mark that's the case you don't know it Google it it's out there right manage your risk put it together the tools are there get

them out there now thank you Alan I mean this is exactly what you described is going through everybody's mind here right so in terms of building trust so AI can scale exactly what every enterprise is

now trying to figure out and what all the things you solving with us has been a great way for us to co-inovate and build the products which will help everybody here as well. So thank you for sharing that. So it brings us to the

sharing that. So it brings us to the next critical piece of our blueprint.

Now everyone please welcome my colleagues John Ifini and Tarun to the stage and thank you Alan. [music]

[music] Thanks so much Amit. It's such an honor to be here in front of you with two absolute legends in the cyber security industry. Please help me in welcoming

industry. Please help me in welcoming Yavghani Drav the co-founder and CEO of Amis [applause] and Turun Fakur the co-founder and CEO

of Vasa. [applause]

of Vasa. [applause] John, thank you so much. It's incredible

to join forces. Let's go.

Thank you so much, John. It's so great to be here. We are super excited to join the service now team. I can tell you having spent most of my life on cyber security there is no better time than

ever to be in this market and the AI is the best tailwind for this market. Let's

go.

[applause] And the good news is securing the agentic powered business uh doesn't start from zero. We're building on a

fantastic base. Our CMDB is already the

fantastic base. Our CMDB is already the IT asset source of data gravity for our customers. Every connected model of the

customers. Every connected model of the enterprise, every system, every dependency, every workflow is mapped in this masterclass powered by our unique

data model. What we're doing now is

data model. What we're doing now is building on that foundation. We've

assembled the right building blocks necessary for security and risk and business outcomes for the agentic era.

First, our cyber asset graph powered by Armis, our access graph powered by Vasa, and the longstanding incumbent knowledge

graph that's been the foundation of Service Now for so long in combination becomes the architectural building block for the security and risk capabilities that the Agentic Enterprise is

unlocking. And all of this is being made

unlocking. And all of this is being made available to our customers and partners through the context engine. Let's start

with cyber assets. The attack surface today has expanded so much beyond it.

It's about OT and medical devices and IoT devices, cloud and code, AI agents.

The attack surface is exploding so much.

At Armis, we are protecting today 7 billion devices and monitoring their day in the life. And imagine that all the legacy tools cannot see and protect 80%

of these devices. Only we can do that.

Absolutely. And AI extends it all to the physical world and expands the attack surface even much more.

That's right, Yenni. And that's why ARMS is so important closing this asset visibility gap even expanding to pre pre-ompilation assets like code. This is

an absolute masterclass that we're assembling right and at armies we are providing continuous visibility and control over any connected device so organizations

can really defend and manage their whole attack surface in every environment every critical area in every industry and now that we have the full picture of

both IT assets and the cyber manifestation of those assets in place what's the next logical access where do we have a not only a

right but a responsibility to play it's access to those assets so organizations get visibility into who and what have access to the what don't you tell us

more about this John thank you this takes me back to July 4th 6 a.m. uh you

know where we deeply had this uh intuition and conviction since the founding of our company which is permissions defined is the purest form of identity and we were built for this

moment of AI and AI you know with AI is being deployed with broad system access commands calling agents talking to agents calling APIs making decisions on

behalf of a human with no human in the loop every L large language model every co-pilot every agent needs an identity and a permission and those permissions are system specific

permissions and the challenge and I've lived this myself for 26 years in cyber is most organizations have no idea what data level permissions are actually allowed and it isn't that you're doing

anything wrong. The incumbent systems

anything wrong. The incumbent systems for previous errors were simply not designed for this error and Vaser is our identity security solution for the

agentic error. Absolutely John and you

agentic error. Absolutely John and you know look we we we um the core data model which we love in this audience uh is what we built with vasa access graph

that John alluded to build before and Amit talked about it earlier over the last seven years with thanks to many many of you in the audience we have

about 120 billion fine grain permissions for humans for nonhumans and for AI agents those identities in the way access graph thank you so much for all

your trust Every agent gets scope permissions with our access graph and lease privilege enforcement at the moment of action. And

these assets and access capabilities are already working together. This is not science fiction. Our vulnerability

science fiction. Our vulnerability response AI specialist which we announced here at knowledge runs an autonomous penetration test over both

pre-ompile and postcompile assets powered by AWS's security agent. Thank

you for your partnership AWS. Fantastic

example of collective defense in action.

This agent cross references its findings with your actual business context. That

third axis I talked about. It then

determines which systems matter the most, who and what has access to those systems, and as a result, what's exposed. In seconds, it turns a list of

exposed. In seconds, it turns a list of vulnerabilities into a prioritized action plan powered by our context engine. The magic is this context engine

engine. The magic is this context engine can feed our workloads but could also feed third party workloads. We truly

want to collectively defend our way of life by meeting you the customer where you are. And then it proceeds with

you are. And then it proceeds with proactive remediation recommendations in real time and that and at scale. Let's

talk about that third dimension which is knowledge. is the intelligence layer

knowledge. is the intelligence layer that connects relationship between policies, vulnerabilities, attack paths and everything across the whole enterprise. This is about making sure

enterprise. This is about making sure that what should be happening is actually happening and we are controlling it. You've already seen what

controlling it. You've already seen what context engine can do and in security it takes on a very special and additional role.

Absolutely. and asset intelligence and access intelligence working in combination form the spine of the context engine for security and risk outcomes. So when Armis discovers a

outcomes. So when Armis discovers a vulnerability on a previously unmanaged device, context engine already knows which production line that asset powers,

which team owns that asset, and what the business impact of downtime to that asset would be in order to remediate findings. And AI control tower sits on

findings. And AI control tower sits on top of all of this. It discovers every AI system across the enterprise. And if

a workload drifts, control tower will catch it and will trigger an immediate remediation workflow.

And with that, you know, we all know in the world of cyber architecture matters.

And with that mindset, you know, we have built very specifically for architecture to be open, especially in the world of cyber where the decision context is available now through service now action

fabric as you heard from Amit yesterday and the platform gets more stronger and stronger once we bring more workloads, as we bring more ecosystem, as we bring other ecosystem partners and build

together with it.

And as part of this responsibility, we just launched the I center for cyber defense to bring together governments, infrastructure providers, and the whole security community to further advance

cyber defense in the AI era. Now, let me give you the bottom line. Service Now

provides the only end-to-end AI native platform for autonomous security. The

only one. And we are purpose-built for aentic business. And we are uniquely

aentic business. And we are uniquely positioned to build the biggest cyber security platform in the world. Any

asset, any identity, any workflow. And

we are the only ones in the world that can do that. Let's go and win, guys.

Awesome. Awesome.

[applause] And John, you know, it's all about at the end of the day product. So, why

don't you and Britney please help how this all comes together?

It would be an honor, man. It would be an honor. Thank you, guys. Thank you so

an honor. Thank you, guys. Thank you so Thank you John [applause] yesterday.

Great job.

Very good. Yeah. Yeah. Nicely done.

Nicely done. Good job. receive the

knowledge from our master class but now suddenly I've morphed into becoming a CVS health uh security administrator.

So CVS Health powers all of the workloads increasingly with AI. Some are

assistive, some automate and some are autonomous. But there's hundreds

autonomous. But there's hundreds literally of models, agents, tools, systems of all types that power CVS's operations today. The moment I log into

operations today. The moment I log into AI control tower in the morning, I see, wow, the AI asset security score has fallen overnight. Why? AI control tower

fallen overnight. Why? AI control tower gives me that answer. A new alert has flagged that the Etna, and for those of you who aren't American, Etnner is a

wholly owned subsidiary of CVS Health.

So, the Etna benefits AI agent has been flagged for anomalous privileges. Wow.

Why does this agent have anomalous privileges? And bear in mind, this agent

privileges? And bear in mind, this agent is an ergonomic workload that helps a member gain insight about their coverage conversationally in their own natural

language. So, the objective of the agent

language. So, the objective of the agent makes sense, but it's at risk risk even of leaking PII. member names, addresses, and phone numbers potentially can be

exfiltrated by folks who don't have the scope permissions for that data. Now, to

be clear, this agent was built with good intentions, but Veronis, another third party partner in the room, thank you for your partnership, detects that it's gained elevated permissions and can

inadvertently share PII with other members, requiring immediate action to stop a potential data leak. In addition,

our armies powered asset intelligence as you can see uh shows how this information can be shared with other systems and other downstream dependent agents. What I've done now is

agents. What I've done now is temporarily disable this agent using Service Now Vaser. I remove elevated permissions that the agent

self-provisioned itself to complete the intent by which it was designed uh to execute. And then as a followup, an

execute. And then as a followup, an exposure record is automatically generated in service now exposure management now supercharged by Armis.

But wait, I found an anomalous uh operational circumstance in this agent.

So now my worry is what else is out there? So the shadow AI detection

there? So the shadow AI detection capabilities of AI control towers tower finds three other agents that were

previously not discovered or managed. AI

control tower brings these agents under management and four new cases are open as a result of its findings. One for the Etna app uh

agent that I previously discovered and three for the incremental agents that were discovered through shadow AI discovery. I mark them all as managed.

discovery. I mark them all as managed.

This triggers continuous security governance and continuous control monitoring across every managed asset, not just agents, but tools, MCP servers,

and all of the other architectural prerequisites that power the agentic error. I want to wrap up with what I've

error. I want to wrap up with what I've shown just to make sure it's clean and piffy. First, we caught an agent with

piffy. First, we caught an agent with elevated permissions that it self-provisioned.

Second, vulnerabilities were discovered from this agent and our early warning workload that we just acquired from Armis helped to enrich the control tower with the incremental signal to capture

that. We executed end to end containment

that. We executed end to end containment with a human in the loop. We um we uncovered shadow AI agents that we previously didn't know about. Um and I have three fingers up instead of four.

and [clears throat] we then created exposure management incidents for all the findings uh with audit ready documentation all an inherent part of the AI control tower. It's important to

note that in this use case CVS health's current modality for autonomy requires an a human to be in the loop but AI control tower can also operate fully

autonomously from monitoring to anomaly identification to remediation. Back to

you Amit.

[applause] [music] [music] All right. Thank you, John, if Tarun and

All right. Thank you, John, if Tarun and Britney. So, when you put all of this

Britney. So, when you put all of this together, AI that senses, decides, acts, and secures across your business, it redefineses how the world works. And our

customers are seeing incredible results.

Adobe is resolving outages 25% faster with autonomous IT. Bell's fully

automated 90% of dispatch related task with autonomous CRM. And Seammens is resolving 210,000 tickets per month automatically with autonomous employee

experience. These are real organizations

experience. These are real organizations driving real results today.

Now, I'd like to move the spotlight from our customers to people we started with today, the builders. The people who take incredible tools and turn them into enterprise ready solutions. Which is why

I'm so excited to have CJ join us, a 12-year community member and 5-year MVP on stage. Welcome, CJ.

on stage. Welcome, CJ.

[music] How are you CJ?

Thank you for joining us today.

Absolutely. Thank you for having me.

Yeah, I'd love to hear from you. How is

your experience with Service Now today different than a few years ago and maybe also talk about what are you most excited about next?

Yeah, absolutely. Uh so first I want to say just how grateful I am to be a part of a community that makes knowledge feel more like a family reunion than a business conference. Right. It's just me

business conference. Right. It's just me and 20,000 of my closest friends.

[applause] So what's different for me is really everything on the platform and specifically though the tooling. Uh so

now we write code in studio, we build apps with uh deploy uh with build agent and we deploy them with release ops, right? None of that was available when I

right? None of that was available when I started. So this is a lot like waking up

started. So this is a lot like waking up in the future for me.

Thank you. Thank you CJ for joining it.

Absolutely. Yeah. And so you know the things that I look forward to I think on the platform uh service now university actually um so for me service now is all about the people right and uh the people

and the community and service now university combines education with inspiration to create a lot of opportunity for people who wouldn't otherwise have it. So the the people are

the next generation of the community and that's why I look forward to service university the most. I look forward to you kind of continue to participate with all of us and sharing all your knowledge as well. So, thank you for joining us.

as well. So, thank you for joining us.

Thank you very much.

Thank you, CJ.

Thank you, Amed.

Yeah, thank you. [applause]

So, before I close, I want to take a moment to thank everyone who made this event possible. Our customers, our

event possible. Our customers, our partners, the analysts, the media who hold us accountable, and everyone at Service Now, and every single person in

this room. Thank you.

this room. Thank you.

[applause] As CJ just mentioned, knowledge exists because of this community and because of all of you here. The last two days have been amazing. We introduced incredible

been amazing. We introduced incredible new capabilities to help you drive more value. Service Now auto is a brand new

value. Service Now auto is a brand new way to get work done end to end with an AI conversational experience across the entire platform. We have 20 new AI

entire platform. We have 20 new AI specialists that execute complete workflows in seconds with even more coming soon. With AI control tower, you

coming soon. With AI control tower, you get live visibility and control of agentic systems at enterprise scale all with a new conversational experience and so many more exciting other things we

announced today as well. So all of this combined means you get the best agentic platform for putting AI to work. So

here's my ask. Head to the expo to explore the AI control tower firsthand.

And don't leave knowledge and go back to the way things were. Take one agentic use case, put it into production, prove it that works, and then expand. And if

you're a leader or builder asking how do we scale AI without losing control, download our blueprint for building an agentic business. It's a deep dive into

agentic business. It's a deep dive into the architecture that makes it real. So

the agentic business starts now. And it

starts with every person in this room.

Let's build, secure, and govern it together. Thank you so much for coming

together. Thank you so much for coming and let's go.

[music] [music]