00:00 Hi everybody, welcome to DEMO, the show where companies come in and show us their latest products and services. Today, I’m joined by Dorit Silbershot, Vice President of Product Management for AI at ServiceNow. Welcome to the show, Dorit.
00:10
Thank you for having me, Keith.
00:12
ServiceNow is a pretty well-known company, right? You’re a big player in the IT service management space, correct? What are you here to show us today?
00:19
All of our customers — and really, companies in general — care a lot about productivity. A recent IDC report showed that IT spending is actually three times more than the U.S. GDP this year. So we know everyone is focused on productivity and making sure they get the best ROI from their investments in the ServiceNow platform.
I’m here to show you our Xanadu release, which came out in September. We’re very excited about this — it’s our biggest AI release ever in the history of ServiceNow. I’m going to show you how, with generative AI and the integration of all these capabilities into our platform, our customers can see significant productivity gains.
01:05
Within a large enterprise, who are these products primarily geared toward?
01:11
That’s interesting because the answer is: everyone. Whether you’re a support agent, a support manager, a knowledge manager, or an administrator of the service, everyone in the enterprise benefits from these capabilities.
01:27
What specific problems are you solving? You must be addressing a lot of different challenges, but with this new release, what’s at the top of that list?
01:36
Today, I’m going to show you how we streamline service delivery across both IT service and customer support. I’ll also show how we help support agents and employees have a much better experience.
01:50
Without this, what would companies be doing? I’m guessing they’d be struggling a bit, right?
01:54
They would definitely be struggling — lots of suffering. They’d also be spending significantly more time working on these issues.
01:59
The goal here is to optimize their time and make companies more productive. I saw some of this before the show — there’s some really cool stuff here. Let’s jump right into the demo.
02:09
Perfect. I’ll start by showing you the support agent workspace. This is where an agent works on a case. In this example, you can see some back and forth between Amy, who opened the ticket, and John.
Now John needs to update Amy, and he needs to write an email. But in the era of generative AI, why should John spend time writing emails manually? He can just ask Now Assist to generate the email for him.
Now Assist can look at all prior communications and all related knowledge articles in the system. It drafts the email for him, ensuring the actions John wants from Amy are well documented. This isn’t just a time saver — it makes John a better support agent.
03:05
It’s pulling data from previous tickets and the knowledge base?
03:10
Exactly. It generates a thorough email with all the relevant information, streamlining communication between the agent and the customer or employee. I can then send the email immediately.
This is just one Now Assist capability that we shipped as part of Xanadu. But we wanted to go even further — we wanted to let customers build their own generative AI capabilities tailored to their specific needs.
That’s why we’re introducing the Now Assist Skill Kit. With a simple prompt, customers can build these skills within ServiceNow.
For example, after sending an email, maybe I want a quick summary of it. That’s not something we ship out of the box, but as a ServiceNow customer, I can easily create that with a prompt. I can evaluate it, turn it on in the UI, and now the email summary button appears. With one click, I get a full summary of all the emails in the case.
If I’m new to the ticket or want to know what’s happened previously with Amy, I just click a button, and all the information is right there. Amazing, right?
We also wanted to support managers, so if I’m an IT service manager wanting to understand team performance, I have this dashboard showing active incidents. But what if I want to see my team’s KPIs today?
Currently, I’d have to work with a ServiceNow admin to create a custom data visualization, but not anymore. In the latest release, we democratized data access. Now, I can create my own visualizations using simple English commands like “Show me incidents by channel as a pie chart.” The system understands and generates the visualization instantly.
06:00
That’s great.
06:03
We’ve also taken it to the next level by introducing AI Agents in Xanadu. AI Agents go beyond generative AI skills — they can autonomously perform work. They think, make decisions, and act on behalf of users.
We’re starting with support agents. Let me show you a real-life story (with names changed, of course).
One of our customers handles a lot of requests around gift cards — extending them, replacing them, etc. Right now, this process requires the agent to search through multiple knowledge articles and figure out what the customer is entitled to. This takes anywhere from two hours to two days.
With AI Agents, that changes. Here’s an example: Matt submits an email request to get a refund for an expired prepaid card. Beth, the agent, receives the request.
Previously, Beth would have to dig through knowledge articles and figure everything out herself. Now, when she opens the case, a notification tells her that Now Assist knows how to solve the issue.
She can view the plan in the Now Assist panel, which lays out all the steps. Beth can challenge the AI Agent if needed — she might ask, “How will you validate if the card can be replaced?”
Since explainability and building trust are critical in the enterprise, the AI Agent explains exactly how it made the decision and how it plans to validate the case.
Once Beth feels confident, she gives the green light, and the AI Agent executes each step. It queries the database, confirms Matt is eligible, and discovers that as a Florida resident, Matt qualifies for an exception.
At every step, Beth provides approvals.
09:27
I like how the agent doesn’t just go from A to Z — it checks in after each step.
09:37
Exactly. That’s critical in the enterprise. The AI Agent also documents its work for full traceability.
Finally, step three is communication. Remember that email generation feature I showed earlier? This builds on top of that. The AI Agent drafts the email, Beth reviews and approves it, and it’s done.
What used to take two hours to two days now takes two minutes.
10:21
Wow — that’s impressive. Is this ready to go now?
10:26
It’s in limited availability right now. We’re working with a few customers who already have it, and we’re aiming for general availability in early 2025.
10:36
I know there are lots of other AI features too. Where can people learn more about Xanadu?
10:46
You can visit servicenow.com for all the details. You can also check out docs.servicenow.com for a deep dive into the Xanadu release.
10:56
Is this all integrated into existing ServiceNow platforms, or is it a stand-alone product people can try out?
11:02
It’s fully integrated into the Now Assist product, which applies to ITSM, CSM, HR, and more. All of these products benefit from these capabilities.
11:13
All right — Dorit Silbershot, thanks again for joining us and showing the demo.
11:17
Thank you for having me.
11:18
That’s all the time we have for today’s episode. Be sure to like the video, subscribe to our channel, and share your thoughts below. Join us every week for new episodes of DEMO. I’m Keith Shaw — thanks for watching.
