ServiceNow Knowledge 2025 Recap
- Paul Nguyen
- May 15
- 4 min read
With Knowledge 2025 now wrapped up, wanted to share some thoughts after attending it last week in Las Vegas.
AI Continues to be the Focus
AI continues to be the focus of the conference, just like it was at Knowledge 2024. This year's focus was on AI Agents. They had a Build an AI Agent lab that in the session planner was listed as 45-minute sessions one after another pretty much every day of the conference.
It was a little confusing since it read like a typical session in the agenda planner you signed up for and in my case, all the sessions were full so I couldn't sign up for any of the available session options. It did say "First come first served" and since this session was on the Expo floor, I went over to it where they told people they could just get in line as they were letting 25 people in at a time.
So some people who had "scheduled" a session were getting out of line since it was taking longer than the allocated time of their session. In my case, I waited about 30 minutes at which point they led us to a small room where they talked about what AI agents ServiceNow has built so far for and was planning on its roadmap for about 10 minutes before then leading us to an open area with laptops setup for us to do a 20-minute lab.
This lab was like other labs at Knowledge where you get an instance through the link provided and had a gitbook to follow along. The main difference was they had laptops set up for you (versus using your own laptop) and the lab was pretty short compared to the usual 100-minute labs as they wanted it to be available to people of all skillsets.
This one worked without issue and for doing this lab you got a free t-shirt. I went to a few other labs where you were using Now Assist to help in your development, using Now Assist to describe what you wanted to do in general layman terms and having Now Assist generate it for you.
For example, creating a scoped app or creating a flow to show how it can speed up the development time. It makes sense why they would want to do this since it not only makes development more efficient for coders but also helps those who aren't as technical be able to build things out.
But as with many other new cutting-edge tech, there were some issues during the labs. In both of my labs related to using Now Assist to create an app and to create workflows (flows, and subflows), the Now Assist ran into issues and gave the "Please try again later" message in the middle of the process as you told it what to do and then answered its prompts for Now Assist to get more details and create the proper app/workflow.
So all the prompt answers and info I provided ended up not being used and after trying 4-5 times, I just eventually gave up and the instructor manually walked us through what it would look it.
What they're trying to do makes sense, just with it being new will be good to see how customers use it and accept its limitations/issues now and be willing to work around it versus those who will stop using it and wait for it to be more fleshed out.
ServiceNow University
Launced at Knowledge 2025 was ServiceNow University, a refreshed way of how ServiceNow is doing its Now Learning program.
They had a big section of the Expo hall dedicated to ServiceNow University where you were able to learn how the program worked now and were able to generate an AI avatar to attache to your badge:

At the booth you could see your ServiceNow Rank which would also update to your ServiceNow University profile (in my case I'm at Challenger Level 3).
They showed at the booth how you were able to build training plans for both you and your team and having AI paths like AI Essentials. One interesting thing was an interactive kiosk where you could see how ServiceNow views as global trends in learning and how AI affects different roles, with different roles changing over time (like how business analysts become more data scientists) as AI becomes more prevalent.
Other than that I didn't see much different from the current Now Learning (biggest thing I did was sign for the AI Essentials path) so we'll see how this improves engagement.
Other Tidbits
Attendance was pretty strong this year with around 25,000, which might even be the highest attended Knowledge ever, but the biggest thing was it felt bustling and the atmosphere and energy of the last Knowledge right before COVID hit.

Even if sessions showed as full, going to them to be on the waitlist was a good option and every session I wanted to go to where I was on the waitlist I was able to get into. So for those who want to go to sessions next year, I would show up to the session and stand outside on the waitlist as more often than not you'll be able to get in since many people who register don't show up.
The Expo hall definitely had more of a ServiceNow feel to it as all their booths were pretty prominently displayed (CreatorCon space, ServiceNow University, the different areas for the Expo panels/sessions, AI Agents, etc). The sponsor booths tended to be more blended in and to the side so you could easily miss/skip them which I ended up doing and only visiting a few sponsor booths.
The closing party being at the Sphere was great. Even though the food was more limited compared to other closing parties that tended to have bigger spreads, being able to go in the Sphere and really seeing how events happen there with the big dome screen is definitely an experience:

Knowledge happens again around the same time next year (May 5-7, 2026) so let's see where things are at this time next year, no doubt in my mind it'll still be AI-focused.