Agentic artificial intelligence promises to fundamentally change the way enterprises operate. Intelligent systems that automate and execute tasks on humans’ behalf, AI agents “will make work easier, faster and more effective, whether by handling tedious manual tasks, speeding up handovers between teams or accelerating market delivery,” the World Economic Forum reported in 2025.
Henry Iversen is the co-founder and chief commercial officer at Boost.ai, a company that delivers AI-powered customer experience solutions for regulated industries. He and hist company are helping organizations prepare for the agentic-AI future.
“With agentic AI, now you can see more use cases around actually executing,” Iversen told The Forecast. “It could … recommend how you actually want to do that kind of mortgage over time, and what is the recommended approach to it.”
In this Tech Barometer podcast segment, part of a series on AI industry leaders, Iversen explains how today’s standard generative AI can help with many aspects of life, like applying for a home mortgage, but agentic AI can go even further. He tells how agentic AI will reshape customer service, and what companies need to do now to get ready.
Managing the Customer Service AI Agent
To drive business results with AI agents, Iversen said customer-focused enterprises must be prepared to move fast.
“Things are moving very quickly, even on a weekly basis,” he said. “There are new models coming out. New Large Language Models (LLMs) are performing better.”
How will agentic AI access new LLMs? And how will companies keep enterprise and customer data private?
Those are just a couple of the many questions Iversen is pondering as he helps end users understand the business implications of AI agents, including the customer service AI agent of the future.
“If you look at contact centers, for example, they’ve been selling seats,” explained Iversen.
He said contact centers are now pivoting to new focal points with customers and prospects. Instead of emphasizing how many humans they have working in the customer service center, sales reps are focusing on how many interactions their AI software can handle. Therefore, contact centers’ economic models may need to change.
“One of the challenges now is just how you put together the future of your contact center or customer engagement, at scale, with all these moving parts,” continued Iversen.
He sees scalability as a looming issue in the emerging agentic AI environment.
“[Because] you’re probably going to have a lot of these AI agents … you want to make sure that you have some level of control,” Iversen said.
At the same time, “you also need to make sure that there is a level of autonomy. If you have hundreds of AI agents in the same way as humans, you’re not going to be able to have control on every action they do.”
With that in mind, organizations will need to manage agentic AI in much the same way that they manage their human resources. With both humans and AI agents, Iversen pointed out, there will be “a lot of similarities on how you train, what information you make available, how you onboard … and also how do you make sure that you have those appropriate guardrails in place.”
Investing in Infrastructure
CIOs must be proactive as they look to bring agentic AI to life, according to Iversen, who said cloud capabilities, in particular, will be a key requirement.
“It needs to be available in the cloud, because a lot of these generative systems obviously need to have access to information quickly,” he said.
Contact centers are among the most obvious use cases for agentic AI. Their IT leaders will have to ensure that applications connect seamlessly to a cloud platform to deliver the needed speed and scalability.
“If you need to be on-prem, you’re going to have a lot of latency and delays. That will not work,” Iversen said.
The technical details of agentic AI implementation remain a moving target, with methodologies evolving rapidly. While IT leaders need to monitor emerging trends, Iversen said what’s most important is getting the basics right. That means ensuring that agentic AI aligns to business strategy.
To that end, Iversen recommends asking questions like: How do we envision the future of customer service?
“Then, start to converse with engineering about what you need to do to get there,” he said. “We need to realize that some things are going to change along the way, and that’s OK. But it’s super important to have this North Star where we know where we are headed.”
Once the journey begins, the user interface should be a major priority, according to Iversen. As AI agents come to the fore, “you should definitely make sure that you’re not spread too thin, that you actually double down on one interface,” he said.
He said the best user interfaces are intuitive for the end user and multi-modal, with the capacity to shift conversations to a human when needed. Here, too, infrastructure is a key enabler.
“UI is definitely going to be the way you win,” said Iversen.
He believes good user interfaces are especially powerful when they’re paired with strong data access. When those two things work together, it’s easy for organizations to recommend the products and services customers need to those engaging with the customer service function.
“You’ll have an excellent upsell opportunity, which is also going to drive revenue for you as an organization.”
Maintaining the Human Element
Even as organizations adopt an agentic AI mindset, customer service leaders must continue focusing on the needs of their human workers.
“There is no doubt that AI will automate a lot of the incoming inquiries in the future,” explained Iversen, who said the rise of agentic AI will require “finding the right balance of things you need to route through to humans.”
When chatbots started fielding simple inquiries, “we gave all the complex stuff to humans, and we thought that was a good thing,” continued Iversen.
He said the assumption was that humans working on complex problems would be less bored and more engaged.
“But what we learned is that actually there was some burnout in some of the people on the customer service side, because with all that complex stuff you need to use your brain extensively.”
As agentic AI comes into play, it will be worth the effort to find the right mix of human and machine responsibilities. For their human workers, Iversen concluded, organizations need to determine how to make the workday “fun, better and also more rewarding — because we know that when people are happy, then they will also give better service.”
Podcast transcript:
Henry Iversen: With agentic AI now, you can see much more use cases around actually executing, not only understand a question and react to that, but also look at the sequence of things and help the customers or employees to make sure that they get an amazing experience. AI technology can open up opportunity to actually have one-to-one relationship with the customers at scale.
Jason Lopez: Henry Iverson is a co-founder of Boost.ai. The company builds conversational and agentic AI platforms that let organizations automate customer services through human-like dialogue. This is the Tech Barometer podcast produced by The Forecast. I'm Jason Lopez. As artificial intelligence continues to evolve, agentic AI is on the road to genuinely reason and do complex tasks. Iverson, who The Forecast reached out to for his AI thought leadership, says this shift could redefine how companies implementing AI get the right balance of people leadership and technology. But it's not that simple. It requires coordination between executives, technologists, and business teams to align on what AI should actually do. And one thing it should do, he says, is bring different apps and tech together.
Henry Iversen: How can we consolidate them? How can we maybe remove some of them? But I think with the new AI technology, it represents the opportunity to have a new interface. And you should get the same experience. I think the multimodality is also extremely important. I want to have an opportunity to talk to my organization, my local bank, my local insurance company, and then continue the conversation on phone, on chat, or whatever, where I need to.
Jason Lopez: The picture that comes to mind is the anecdote of changing a tire while the car is still moving. Iversen says the biggest challenge is things are moving very quickly.
Henry Iversen: Even on a weekly basis, there's new models coming out, new large language models performing better and stuff like that. So there's obviously a lot of things going on in this space, which again also have kind of implication in terms of like, when do we get access to the model? What is the privacy around that? How do we need to think about like actually implementing this in my own kind of call center? So there's a lot of things which is also outside the bigger ecosystems.
Jason Lopez: With new language models emerging almost every week, the technology is evolving faster than most organizations can adapt. And that constant change doesn't affect just the tools themselves. It reshapes how companies think about scale, privacy, and the entire structure of their customer operations.
Henry Iversen: AI can handle thousands or millions of interactions. One of the challenges now is just how do you put together kind of the future of your contact center or customer engagement at scale with all the moving parts?
Jason Lopez: Perhaps a deeper question is how the organization works with AI. It's a given. No one is under the illusion that AI just magically works. It takes effort. And that begins with getting executives, IT, and business teams on the same page to implement and use it.
Henry Iversen: Being able to really align the organization in terms of what is the needs of the future is pretty hard, I think. What I've seen so far is just make sure you can have those alignment early on with the executive leadership, but also the tech and the CIO also involved. Because everyone has different perspectives, but you really need to have that alignment. I think another key consideration you need to make is also, okay, are we going to operate and manage these systems over time? Some of these systems already out there is built for technical people. And then how do you scale up with that internally where you maybe want to have a bit more kind of a no-code interface? The business is definitely extremely important, that the business is a key part of this.
Jason Lopez: On the technology side of the equation, the company's cloud infrastructure is vital. Iversen says cloud-based systems are essential to enable AI models to function effectively.
Henry Iversen: For everything we do, it needs to be available in a cloud because a lot of these kind of generative systems obviously needs to have access to information quickly to make sure that everything works between the clouds so you can make sure that you reduce the latency. Then also you need to manage the costs. So having a vendor which can actually manage and make sure that the cost is not taking off is also extremely important for the overall business case for it.
Jason Lopez: But as critical as the IT side is, he says the business vision should drive AI partnerships and architecture. The starting point isn't speeds and feeds, but rather the organization's business goals.
Henry Iversen: How do we envision the future of customer service should look like? You can also then start reverse engineering what you need to do to get there. And we also need to realize that some things are going to change along the way, but that's okay. But it's super important to have this kind of North Star where we know where you're all headed. And that also brings a lot of kind of clarity but also alignment into the organization.
Jason Lopez: It's one thing to imagine the future, says Iversen, to map out that North Star and picture customer service. But at some point, vision meets reality. And that's where an important question comes in. Does it make sense from a financial point of view?
Henry Iversen: There's a lot of cool stuff we could do, but again, there's also a cost element to it. But I also think it's just extremely important to make sure that, okay, what level of support and experience are we going to bring out to the market? One overlooked possibility and opportunity with agentic AI is also that you can do something you haven't done before. I have one example actually from a bank where they offer new services because now they can use, say, agentic AI to handle a lot of the workload of offering that service. Some service could be made available from customers, but it's just too time-consuming. Another example could be, I could potentially, if I have one million customers, I could potentially call all of them. But again, that would probably not be very scalable if I'm the CEO of a bank to call every customer. But with AI technology, you can open up an opportunity to actually have that one-to-one relationship with the customers at scale. We can obviously automate existing processes with agentic AI, but for me, it's more like, okay, how can we open new opportunities with agentic AI?
Jason Lopez: And that question of how AI can unlock new opportunities rather than automate the old ones leads to a deeper point. Because innovation isn't only about efficiency, it's also about elevating the experience. As Iverson explains, the real challenge is to use these new tools not just to do things faster, but to do them better for customers and for the people who serve them. Once AI agents start acting with a degree of autonomy, the question becomes, how do you guide them responsibly without holding them back?
Henry Iversen: Yeah, that's a good question. And I think one element which is extremely important as you start scaling these systems is that you're going to have a lot of these AI agents. You want to make sure that you have the level of control, but you also need to make sure that there is a level of autonomy into it. Because you can imagine if you have hundreds of AI agents in the same way as you have humans, you're not going to be able to have control on every action they do. So you need to make sure that they can do that within a controlled space, but also being open that, okay, if there's a thing I'm unsure about, then I should maybe ask you for approval. So make sure that you have the human in loop for some of the questions coming in.
Jason Lopez: Finding that balance between control and autonomy brings the human into focus. As AI takes on mundane tasks, the role of people becomes less about managing volume and more about creating meaning. It's a shift Iversen says will be essential to how organizations design the future of their customer experience.
Henry Iversen: There's always going to be customers needing to get help from a human. But I also think that back in the days when we started kind of doing all the simple inquiries to chatbots and we gave all the complex stuff to humans, we thought that that was a good thing because all the simple stuff, humans don't want to go answering those things. But what we learned is there was some burnout in some of the people on the customer service because all the complex stuff was obviously, you need to use your brain extensively to kind of solve those. So I think just finding the right balance of things you need to kind of route through to humans is going to be pretty essential to make sure that they also support customers.
Jason Lopez: Henry Iversen is co-founder of Boost AI, a company that was founded in Norway. It builds conversational and agentic AI platforms. This is the Tech Barometer podcast produced by The Forecast. I'm Jason Lopez. You can check out more technology stories. We have them at theforecastbynutanix.com. That's The Forecast webpage. Again, that's theforceastbynutanix, all one word, dot com.
Adam Stone is a journalist with more than 20 years of experience covering technology trends in the public and private sectors.
The podcast was produced by Jason Lopez, executive producer of Tech Barometer, the podcast outlet for The Forecast. He’s the founder of Connected Social Media. Previously, he was executive producer at PodTech and a reporter at NPR.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.
Related Articles
