Maybe it is because I am not a full fledged developer and AI has basically allowed me to stop learning how to script but I feel like I have gotten better responses and solutions from this subreddit than the ServiceNow community. The response time is also another thing, people usually reply right away. ChatGPT tends to hallucinate a lot on how-to's compared to just wanting script. Thoughts?
So I am trying to set up 2 different kinds of email notifications for incidents and req items. One of the conditions that I can choose [Additional comments] [changes]. There is no way to determine who it the person adding the comment meaning I can not do [Additional comments] [changed by] [User].
I understand that additional comments and work notes are not fields of the incident or ritm record and has a seperate entity in sys_journal if I am not mistaken. So right now my notification is set to 'who will receive' containing the assigned_to and is given a link to ${URI_REF} which links to the platform UI . And then a nother notification with same condition but sent to user with a mailscript linked to the service portal ritm record.
So what happens is both notifications fire to both people if either one of them add a comment and they see it in their own respective sites. in other words if user adds comment, they still get an email about a comment being added, same as agent.
Its really not THAT big of a deal but I can see maybe in the future some VIP personnel might find it annoying. Any reliable workaround?
I would think by now this would be super easy. But our shop is still doing "for text" searches agains all incident and all request views.
When searching for such tickets, you're not sure if things were put into a specific category. And sometimes all tickets aren't incidents (we classify how to tickets as requests).
There has to be something we are missing. How do you do it at your shop?
Hi everyone,
I am currently exploring a POC where we want to integrate Microsoft Copilot Cowork with ServiceNow.
Scenario:
A user raises a ServiceNow ticket like:
"VPN not getting connected"
Expected flow:
I have some architecture questions:
Examples:
Looking for guidance from anyone who has worked with Microsoft Copilot + ServiceNow integrations.
Thanks.
Earlier this year, we started a project to review our current ITSM tool and to start looking at other vendors, but not necessarily committing to switching tool, just doing our due diligence. A month or two ago, we got a new CIO who came from another company that used ServiceNow. Within the last 2-3 weeks, things have escalated, and it looks like we're now moving to ServiceNow.
I am currently the administrator of our current tool (just me). As a totally new customer, what advice or warnings do you have for me? Is there anything I should start doing to get out in front of this change?
A Bit of background of me, I have 10 years of experience in ITIL/SIAM/ITSM consulting. Most of my work has been across banking and captive environments, using ServiceNow primarily as the ITSM suite of choice.
Over the next few years, I’m considering moving more toward Product Management, Pre-Sales, or Solution Advisory type roles. At the same time, one thing keeps looping in my head, "You should know scripting." I’ve heard it from managers, architects, consultants, random LinkedIn posts basically everywhere.
Now just to be clear, I have zero resistance to learning JavaScript. In fact, I want to learn it because I feel it gives much more control over the platform and makes you more self-sufficient. My issue is more with how people suggest learning it.
The advice I constantly hear is: "Just watch a few random ServiceNow JavaScript playlists," or "Just jump into platform scripting and you'll pick it up." I’ve genuinely tried that approach and it completely breaks my brain.
I suspect I’m wired differently (possibly ADHD tendencies). My brain hates fragmented learning. I can’t vibe-code or randomly jump into scripts and absorb concepts. I've tried and i simply can't. I’m the opposite. I’d rather go buy 2 to 3 JavaScript books, start with ES5 fundamentals, understand programming basics and web concepts for a few months, and only then touch ServiceNow scripting. Structured learning somehow works much better for me.
Has anyone been able to crack this one?
I wish to use port a react app into ServiceNow however the presentation looks messy with the customer UI page endpoint.
From looking at the manuals i can find no other way to present my app.
Has anyone figured out a way around this?
with the service portal I would just put my angular widgets in a page.
Title
Hello ServiceNow folks, I heard there were some silent layoffs in ServiceNow India today in Customer Outcomes and other team. Anybody aware of this insider news and how many impacted. Sorry for those impacted. 🫤
I have my CAD exam tomorrow and I have completed reading ebook and also completed doing labs. I have some knowledge on ebook and I was going through mock tests from certyIQ pdf and I found out that many answers are mismatching or wrong when compared with other websites like examprepper or chatgpt/Gemini answers.
Could anybody please help me with this, which one should I follow now certyIQ pdf or examprepper or chatgpt answers??
In CMDB, they say there has to be 1 CI for 1 instance.
For example, lets say Azure Databricks.
It can have multiple instances owned by many different teams but there should be only 1 CI - Azure Databricks.
Now, different teams are saying they each need to be change approver group on this CI, so they are asking us to create multiple Azure Databricks entries but we know this goes against general CMDB principles.
So, 1 CI but multiple change approver groups.
How do we resolve this dilemma?
There is Azure Databricks. There are many teams in the org that use it - Team A, Team B, Team C and so on ............
On Azure Databricks the CI itself, we can specify who is the Change Approver Group.
If Team A is specified as the Change Approver Group, and if you submit a change request with CI field populated as Azure Databricks, then change goes to Team A
The problem is, Team B and Team C are saying we want to submit change request for AZure Databricks, but its going to Team A, so , create another CI and make our team Change Approver Group so that we can approve it.
But, creating multiple CI's for Azure Databricks is not a good idea is what the architect told me. We need to have just 1 CI. Also, our change management process owner had told us we won't run into this situation.1 CI will have 1 team as approver is what they said
How do I deal with this?
I have been in the ServiceNow ecosystem since 2013. Their core benefit has always been the ability to unify data and workflows across different areas of IT (in practice) or the entire company (hypothetically) on one platform.
I had assumed that ServiceNow would double down on this concept with their attempts at becoming an AI company. Data in a unified format would be super valuable for AI models to figure out what to do. A unified way to trigger actions would be very valuable for AI agents.
But their actions imply they no longer care about this. They acquire off-platform solutions left and right, do not replatform them, and are actively competing with their own (or their build partner’s) on-platform apps with ServiceNow owned off-platform apps which ServiceNow actively markets.
It seems to me that they spend every single opportunity they get shoveling increasingly platform agnostic AI dreams, and wherever that conflicts with the unified platform model, they ditch the platform concept.
Meanwhile I barely hear them talking about the importance of unifying UX on the platform, the importance of unifying data in the common data model or the importance of a unified architecture and automation engine. I also don’t really see investments in said core platform. Every time I try to ask for new core features I get told about some new AI concept their own sales org can’t explain.
It seems to me that they have not only abandoned the core concept they have built their platform around, but actively undermine it at every opportunity.
Hey Guys,
on friday I have my second chance to pass the CSA exam. First time I was so confident learned a lot with Udemy learn courses but the questions where totally different... sometimes i didnt knew any answer on some questions... do you have any tipps to prepare myself better for it?
tysm for helping me
I made a post about my experience here: https://www.reddit.com/r/servicenow/s/ZvA1tL3N31
I have a niche at my current company as the SAM sme, soon to also be involved in HAM. Lots of experience in Change and Problem management already and am the sme for both.
I assume that I should focus on CIS-DF next. But, I have seen posts that say that CAD is helpful.
The biggest shortcomings as an org we have are building catalog items, data quality, reporting, and just overall platform expertise.
Frankly, I’m looking for as much leverage as possible. Im looking to increase my role at my current company or to move on in the next 10-12 months.
I have 25 years of experience across various areas in IT.. been an agent on the platform for about 3 years. I have been tagged to move into a platform ownership role and sme for most IT functions. One other person has been tagged with me. But they are very passive.
I’m the only CSA in the org currently. We rely heavily (too heavily) on our implementation partner.
Thanks for any advice.
You can run the benchmarks for yourself using the "Run ServiceNow Eval" Action in AppAgent Chrome Extension:
https://chromewebstore.google.com/detail/appagent-for-servicenow/jafoppdjbleekamickhclhdaoephgimi
Note that Opus 4.7 quality changes drastically from one run to another (due to agressive quantization maybe)
First I want to say thank you to this sub. I took a lot of advice that I’ve read here.
My advice for others..
Yes. Read the blueprint and study the e-book.
Take every practice test you can find, especially MeasureUp. I also bought the ones from Skill Cert Pro.
On the exam, take your time. The questions are worded in the most confusing way you can imagine.
Eliminate obvious wrong answers. You usually get down to two that are close to correct. Reason through the last couple. If you know the concept you will know which one is correct.
On subjects where there are lists of items. Memorize them and be prepared for “select x correct” type questions.
Go study the security center outside of the ebook. Know which function you do x.
I think I missed 6 to 8 questions. Nearly all of them were navigation/UI related. I have some platform experience as an agent, but didn’t spend much time in my PDI. You should.
Know business rules, Client Scripts, UI Policy, UI actions, data policies inside and out (from the e book.)
Spend time on tables, where to modify them, relationships, etc.
Also, there’s more Catalog related questions than I expected.
Best of luck to you.
Hey everyone,
We just went through a major multi-release migration, jumping from Yokohama all the way to the Australia release, and we’ve run face-first into an architectural nightmare regarding Yes/No Catalog Variables.
As many of you know, ServiceNow modernized the Flow Designer engine a couple of releases back (around Zurich) so that Yes/No variables natively evaluate as true JavaScript Booleans (true/false) rather than their historical strings ("Yes"/"No").
To mitigate this during our upgrade, we deployed the documented system property:com.glide.hub.flow.catalog_variables.yes_no_as_boolean = false
The Trap:
This property looks like it works at first glance. It successfully forces native Flow Designer data pills (like from Get Catalog Variables) and standard out-of-the-box IF conditions to continue returning and evaluating the legacy strings ("Yes"/"No").
But here is where the property completely abandons you:
1. Flow Designer Custom Actions (Script Steps): When a data pill is passed into a Custom Action input, the runtime engine seemingly ignores the system property. The data enters the JavaScript step as a literal primitive Boolean true or false. Any legacy code checking if (inputs.variable == 'Yes') fails silently because it receives a boolean instead of a string.
2. "Submit Catalog Item Request" Action (Cart APIs): When using flows to dynamically submit other catalog items, the underlying Cart API layer strictly validates incoming payloads. It expects Booleans, ignoring the property entirely. Passing a legacy "Yes" string results in dropped data or payload errors because it's pushing a text string into an endpoint that strictly expects a Boolean.
The Problem:
While we know we can fix this via inline scripting f(x) on action inputs and updating the JavaScript inside custom actions to look for (inputs.variable == 'Yes' || inputs.variable === true), doing this manually across instances introduces massive regression testing risks and heavy manual refactoring.
Has anyone solved this?
If you migrated a footprint from Yokohama/Washington to Zurich/Australia:
• Is there another hidden, global system property that actually forces compliance across all execution layers (including Custom Action script steps and Cart APIs)?
• If no property exists, how did you handle updating these touchpoints? Did you write a fix script to mass-update flow definitions/inputs, or did you honestly have to bite the bullet and manually refactor every single script and mapping touchpoint?
My organization is considering Now Assist, but the price point has them skeptical. Does anyone use the Now Virtual Agent without Now Assist powering it that could give me an idea of how it works for your organization? Any pain points or things to consider? Trying to figure out if it’s a viable option to use for the interim while senior leaders make room in the budget for it.
Thank you for any advice you have!
Hi, I am Anders.
I spent most of my time at Knowledge listening to promises of a great future, getting invested in them, and then simply trying to find out what was true. I spoke with several ServiceNow product people - and for some questions I got great answers, while some replies were vague, and some times I had to give up and make an educated guess based on what I know.
Here are some of my thoughts.
(Disclaimer: I am not writing this on behalf of my employer or any company. My words, ideas and conclusions are my own, I am not claiming to know anything, I am most likely wrong or didn't get the memo, never listen to me or do anything I do, and if you get upset or angry based on whatever I say here I apologize. Just sharing my thoughts, because I think it's good to share ideas.)
Moveworks
Bill McDermott keynote quote: "This is now fully integrated into the platform". At worst this a false statement, at best it is a misleading thing to say.
My investigations show that Moveworks is still an external service, not at all in ServiceNow (at least yet), and is fully running, and is configured and maintained at moveworks.com only. It has a connector to ServiceNow, so the "fully integrated" part is at best an inbound integration from Moveworks servers. All demos you see will either be directly at a Moveworks instance, or an embedded chat (like Virtual Agent), like an iframe type, but you are actually typing in Moveworks and not ServiceNow. Any question I had around how to set it up, adjust prompts, set up custom integrations and so on were for the most part ignored or I was told a roadmap would have great ServiceNow functionality. Educated guess: Besides the connector to ServiceNow (no different than other connectors from there) you are just working in another system, which is in a different datacenter, meaning your CISO might want to have a word in any decision now that company data goes to more places.
I got my hands on a lab instance and made two Youtube videos (part 1 and part 2) where I "unbox" Moveworks, which is where I got my initial questions from.
(Disclaimer: I am not actually talking down Moveworks here. I actually think the service itself seems pretty awesome, it's super fast, and compared to Virtual Agent it's a way better user experience - my critisism is the misleading communication around this being anything but an external service, and how my questions around it are still unanswered or are mostly ignored.)
Autonomous Workforce
I was trying to figure out how this is more than a collection of preconfigured AI Agents, and why the current Agentic Workflows won't already cover this. The idea, as I understand it, is to provide agentic functionality on task level without having to work with complex prompting. This led me to question how much these agents are able to do, but on one level was hoping someone would surprise me with a great idea on how they would learn from historical resolutions.
I filmed asking a representative demoing this in the Expo area, a video which is also on Youtube. Direct quote: "We've tried to make this as plug and play as possible, so you don't have to configure anything". I asked "How does the agent understand HOW to resolve a ticket?" and the response was more or less that it would only do simple things like triaging, looking up knowledge articles, writing comments and resolving. Don't take my word for it - check out the video and make your own conclusion.
He does also say that if an agent would need to perform an integration to solve the task, that's probably something they need to add at a later point. My educated guess, this is pre-prompted agents with a generic problem-solving capability on a very basic level - assigning, replying with self-help text based on articles, and probably not much else. Can you set up tools like you can with AI Agents today? No. Can you direct these "AI specialists" to be more process specific? No.
At least not now.
AI Control Tower
I think the AI Control Tower is a great concept, and much needed now that companies buy AI services en-masse and with that need some level of control. I loved the demos showing a kill switch for rogue AI agents in the enterprise, and that a new role (AI steward) is introduced to be the "firewall persona" for people who wants to place AI everywhere with no risk assessment performed. We need it, and it's a great idea. I am not a fan of the user interface still, but assume it will improve at one point - so not spending more time on that part - it's new, I understand.
What I wished they did spend more time on is talking about just HOW the AICT can govern and control all AI in the company, and with what requirements and prerequisites. I had hoped they would explain early on how you need to perform discovery to find AI components, that you most likely need to use Service Graph Connectors for this, and how this is basically a CMDB monitor on AI classes.
Sure, this is a technical thing, not suited for a keynote, of course, sure thing, blabla.
But as soon as I spoke with people on the customer side right after, and they wanted to get going ASAP, I had to spend quite some time explaining that without spending time and money on that messy CMDB, and having a good quality Discovery process first, this probably will require to start that process before even considering it - and by experience that will take lots of time. What I was hoping for was for ServiceNow to at least mention the CMDB connection at all, so customers can be aware of what they have to do. Yes, they should have fixed their CMDB a long time ago - don't even mention it.
In any case, for this part I think it might be one of the more important features moving forward, it's a great idea, and I think having proper governance of AI in the enterprise is extremely valuable. It will be prosperous times for CMDB-experts the next few years.
Employeeworks / Employee Slate
This could have been under Moveworks, but I am here talking about the new portal they are introducing to support a Moveworks-chat / input field, and it will populate widgets under it. It is most likely a mix between the Moveworks embedded chat window (which integrates back into ServiceNow) and local widgets with the results. I don't have much to say here (my Moveworks point above points at my thoughts around that), other than I am really wishing for this to work out. I don't like the fact that we are stuck with Service Portal, I think it's outdated, and I am hoping for something more scalable and future looking. On it's own I think it's great that new portal tech is introduced, and I hope it will be something we can customize as much as we want (I doubt that though).
They are limiting all of this to employee-side (so IT and HR more or less, to start with) - my educated guess is because of what exists in Moveworks already when it comes to their connectors. Let's hope they make this to work in a broader context, or at least allows us to set things up how we want - and also support end-users / external users. At the moment, that's not happening though. I haven't tried this, only seen it on a screen - as soon as I get a hands-on experience I will make a video about it. Crossing my fingers this will be awesome.
Other things
There was just so much going on. Agent Fabric, Context Engine, Otto (just a brand by the way), Warpspeed, Build Agent, and new company aquisitions. I could write about these things too, but for me personally I need more time to get more information.
And there are things I wished was introduced which wasn't, so I guess I'll just make that myself.
Thank you for attending my TED talk, this was brought to you by not AI.
Hey Everyone, I’m looking for ServiceNow Tech Leads that have in depth experience implementing Discovery (on premise & cloud), Service Mapping (top down or tag based) and event management sources.
The main thing we are looking for here that most are lacking is strong scripting experience. Most of the Architects I’m running into aren’t as hands on as they used to be (which is fair, and not a bad thing as their roles have progressed to different responsibilities) but for this client - they need this person to be able to get in the weeds when necessary.
It’s paying up to $190-210k base, plus equity, plus unlimited PTO, up to $20k in utilization bonuses and a lot more.. fully remote too.
If you know anyone, I’m happy to pay a referral bonus if I place them.
Send me a message if this resonates!