When Pat Casey joined ServiceNow around two decades ago, he was employee number nine at the digital workflow company. It now has than 26,000 employees worldwide.
Casey is credited as one of the founders of the company – although not, he stresses, the founder. That title belongs to American Fred Luddy, who set up ServiceNow in the US in 2003 as Glidesoft.
The pair had worked at Peregrine Systems – Luddy as chief technology officer, Casey as a product author – before the company went bankrupt. After that, Casey moved on to Adobe Systems before joining ServiceNow.
“When I was hired, there were only two job titles in the company: there was developer and Fred Luddy. And I was not Fred Luddy, so I was hands on the keyboard for probably the first 10 years or so,” he says.
READ MORE
“I gradually moved up to into management. We had a gap there. I don’t think we had a lot of people who knew the technology, knew how the teams worked and the personalities, how to actually get the stuff done.”
These days, Casey has multiple job titles. He is chief technology officer, and executive vice-president of development operations, leading a team of more than 9,000 people across AI, product, and quality engineering, developer productivity, cloud services, advanced technologies, and customer service and support.
“I’ve got a really strong team. My phone doesn’t tend to ring at two o’clock in the morning much these days, but it used to,” he says.
“My wife had a coffee pot she would just set out most nights, and she would say ‘Your phone’s going to ring at two o’clock in the morning. Just come down and start the coffee’. About half the nights of the week I would be hitting that coffee pot at 2am.”
Casey has seen ServiceNow navigate some of the best and worst times in the tech world. But the most significant development for ServiceNow was not just the internet but the point at which people figured out how to use it effectively in businesses.
“I’d say it took the industry five to seven years to take that technology from working technology to a family of use cases, and that’s sort of where ServiceNow came in. We’re like, ‘Hey, we can take this internet technology, we can apply it to make business workflows better’,” he says.
“If we had tried that in 1995 we would have failed, because the technology wasn’t there for it, whereas 2004 it had just reached the point where you could pull it off.”
Since then, there have numerous developments – cloud computing, the proliferation of data centres, and now artificial intelligence.
That has translated into a business that generated more than $3 billion in revenue in the first quarter of 2025, with a growth rate of around 19 per cent year on year.
In 2024, it generated more than $10 billion (€8.56 billion) in revenue, and recorded gross profit of around $9 billion. Again, growth was in double digits compared to the previous year.
ServiceNow is already utilising the new generation of technology, integrating AI-based large language models into the company’s products. Casey says it is still early days for AI though.
[ AI use by Irish companies doublesOpens in new window ]
“We’ve got a technology that is still very rapidly evolving. The models of today are dramatically bigger and more sophisticated. But the use cases – people are still struggling to find a really resonant set of use cases.
“And you see some false starts in the world of consumer AI, a lot of us have AI fatigue,” he says.
While AI is disappointing consumers in some areas, Casey thinks the expectations of AI are more closely aligned with what the technology really does in the enterprise space.
“In the consumer space, you want a personal assistant that is empathetic and understands you. Technology is not there yet. You can mimic aspects of it but I think you end up with sort of this technology uncanny valley – it’s good enough to tempt you to use it, but not necessarily good enough for you to delegate tasks to it,” he explains.
“In the enterprise space, it tends to be more structured interactions. The business world actually works on structure, which is easier for these models to operate over.”
With the technology changing so rapidly, it must be hard for companies to stay abreast of things. Casey is quietly confident that the company can keep up with the current demands of its customers and their future needs. Key to that is making sure the platform isn’t dependent on a single model, and that newer models can be easily added as they are developed.
“We’ve mitigated the risk there, because we have separated our logical layer from the underlying model implementations. We have made a choice that we’re going to support all the big frontier models, and we have internal models as well.”
It is not just about delivering the technology to its customers though. ServiceNow uses AI in its day-to-day business, including in coding. However, Casey does not see a point just yet when AI takes over the entire process, as other tech companies have mooted, with small productivity gains rather than earth-shattering shifts.
“The place where it still falls down is it doesn’t know our code base. So if I asked it to do something sophisticated over a big, beefy code base, like ServiceNow’s it doesn’t really know beyond the class it’s looking at. So that ability is not to my eye there yet for a big enterprise-scale code base.”
Despite recent reports about AI sabotaging instructions that would shut the models down, or threatening blackmail for human workers who may try to shut them down, Casey is not concerned about AI going rogue as such.
Today’s models are less sophisticated than they seem, he says, describing it as the illusion of consciousness.
“There’s no awareness loop in there. It’s just a very sophisticated system that it can act like it understands, but like there is even no concept of shutting down a model. It’s stateless,” he says.
“I’m less concerned about those scenarios with today’s technology. The thing I keep harping on in the enterprise context is people need to think about the security context, where they’re willing to trust and run the AI.
“The assumption that the AI agents are going to be better behaved than humans, I think is a false assumption. From an ethical standpoint, most humans are going do the right thing most of the time.”
One risk he sees from AI though is in the training of new engineers. As AI agents take over more and more tasks, it leaves fewer opportunities for new engineers to cut their teeth and gain the experience they need to progress.
“The challenge would be ... how do I get a smart 18-year-old kid, who’s motivated to work in this field, trained up on algorithms and the codebase?
“Because right now, half of that’s in school, and half of it’s on the job... It feels like we’ve got to change the pipeline there so we still have people flowing into the field.”
It is not just AI that is causing uncertainty in the tech sector. The current geopolitical instability, a potential trade war sparked by the Trump administration, the threat of tariffs and repercussions against companies that do not bring operations back to the US – all are causing ripples of anxiety, which is never good for business stability.
[ Q&A: What does the latest Trump tariff move mean for Ireland?Opens in new window ]
So far though ServiceNow has not seen a big problem. “We won’t say uniquely in the tech industry, but we’re in a weird ‘tweener’ state. We’re obviously US based, so we’re going to follow all the US laws, but we’ve got a very big presence here [Ireland],” he says.
“We’ve got an obligation to follow the legislation and the laws here. So I think the dynamic we’re seeing right now is that we’re in a bit of a wait and see. We clearly would love to continue doing business successfully here, and right now, we believe we can, but you can’t predict the future fully.”
Dublin is one of the company’s biggest non-US engineering sites outside India. Casey says it originally began as user interface development, but pivoted over the years to application and tool development, too, along with AI teams.
ServiceNow is in its growth stages here, with the announcement of 400 jobs in 2023. The original plan was to grow over three or four years, but Casey says they are in advance of the schedule. “The advantage of Dublin is, this has been a big tech centre for 20, 25 years,” he says. “There’s a lot of domestic talent here.”
Is there anything he would have done differently in his career?
“I have made a lot of choices in my career that were career optimising. I look back at this point, like, I’ve been extremely successful, I think by most people’s standards, but I missed out a lot of stuff in my family,” he says. “I wasn’t there for a lot of my son’s basketball games; I wasn’t there for a lot of my daughter’s dance recitals.
“Right now, I can spend a lot of time with them. There’s benefits, right? But there’s costs, too. There’s a cost benefit there. I don’t know if I’ll be lying on my deathbed thinking ‘Man, I was on the right side of the cost-benefit curve’, or if I may be laying on my deathbed saying, ‘Man, I made the wrong call’. I just know this is where I landed right now.
“I’m an engineer, I like building things. I personally take an enormous amount of satisfaction out of building stuff and seeing people use it. To me, that’s the loop that keeps me going. That’s the thing I couldn’t I don’t think I could operate without.
“If there was a way for me to get that same feedback loop with less side effects, I probably would make that set of choices. But right now, I don’t know that there is, or there was.”
CV
Name: Pat Casey
Job: CTO and EVP, Dev Ops, ServiceNow
Lives: Del Mar, California
Family: Corinna (wife), Allison (daughter), Anthony (son), Whiskey and Saki (dogs)
Something you might expect: He has embraced AI and remains realistic about its benefits.
Something that might surprise: He was once chased out of the woods by a black bear while backpacking in the Appalachian Mountains.
