Using agentic AI to work smarter, not harder

In a recent interview with PYMNTS, Karen Stroup, Chief Digital Officer of WEX, offered a fresh perspective on how agentic AI is helping businesses move faster, work smarter, and unlock new efficiencies. Her key message? Businesses that are brave enough to experiment with AI, especially agentic AI, are the ones that will thrive today.

So what is agentic AI, and why does it matter? Let’s break it down.

What is agentic AI?

Agentic AI is a next-generation form of artificial intelligence that does more than just respond to inputs. It can make decisions, take actions, and even adapt to new situations without needing constant direction. 

Karen says, “There are two key things to agentic AI solutions. The first is the ability to reason and make decisions — taking in new information, learning, adapting, and most importantly making that decision. The second is it can take action. It goes from presenting an insight to taking action for you.” Karen describes this as “game-changing in terms of the opportunities to improve how we work.” Unlike traditional AI, which is task-based and rule-bound, agentic AI has more autonomy and can manage complex processes end-to-end.

Think of it like this: if traditional AI is your assistant, agentic AI is more like a colleague. It can understand goals, adjust to new information, and figure out how to get from A to B without you holding its hand the whole way.

Making work more efficient

One of the biggest opportunities Karen sees with agentic AI is around efficiency. In the interview, she points out that many business processes, especially in industries like payments, finance, and logistics, are still bogged down by manual work or legacy systems. Agentic AI can step in to automate these processes, but in a more intelligent and responsive way than previous tools. And she makes it clear that agentic AI is not a replacement for humans. 

“For me, the key is that agentic AI solutions are not about taking the human out of the loop — it’s about enabling them. It’s about taking those insights and decisions, and inserting the human to review. That’s where the payments world is going. The combination of improved insight and action while maintaining control.”

For example, agentic AI could handle a supplier onboarding process: collecting data, verifying information, resolving small issues, and kicking off payments. Not only does this save time, but it also reduces errors and helps teams focus on more strategic work.

As Karen says, it’s not just about doing things faster, it’s about giving people “a better experience at work” by removing friction from their day-to-day responsibilities.

Learning by experimenting

Agentic AI is uncharted territory for a lot of people. Determining its use, cost-benefit, and whether it has a place in your organization requires some experimentation. Karen says, “If you’re going to experiment with any type of AI solution, you’ll want to focus on two things. One: focus on areas where you’re most likely to see success. And two: make sure there is a good return on that investment?”

Karen emphasizes that the companies making real progress with AI aren’t necessarily the biggest or most tech-savvy. They’re the ones that are willing to experiment. And experimentation doesn’t have to mean taking on risky, complex projects. It can start with something small and well-contained, like automating a single repetitive task.

She explains that the best results often come when businesses pick a process with clear boundaries, strong data, and real business value. From there, they can test and learn while involving legal, compliance, and IT early to make sure everything is safe and well-governed.

This mindset of “learning by doing” is what separates businesses that adapt quickly from those that get left behind.

Balancing trust and autonomy

One challenge with agentic AI is trust. How do you trust a system to make decisions, especially in areas like finance or customer service? 

Karen says, “It starts with a really strong data governance policy. From there, you want to have a process by which you can verify and build confidence along the way. Trust is earned through transparency, traceability, explainability and through real-world validation.” Ultimately, companies need to know how the AI works, what data it uses, and how decisions are made before they can truly build confidence using it.

A smarter way forward

Implementing AI into the workplace isn’t just about the technology. It’s about how people and tech can work better together. It’s a mindset shift. The idea of creating “agentic enterprises,” where AI becomes a partner in everyday operations is not about replacing employees, but rather enabling them to do higher-value work and have a more collaborative relationship with AI.

“What’s exciting right now is that we have new opportunities to solve problems in new ways. The best way to unlock it is by experimenting and learning through secure and frequent iterations to see what’s possible.”

And according to industry leaders, this is just the beginning.

At WEX SPARK Health and Benefits 2025, we heard how teams across WEX are bringing innovation to life, from unified experiences and smarter APIs to early applications of agentic AI. It’s work grounded in purpose, trust, and thoughtful experimentation. 

To see more of this in action, watch the SPARK closing Product and Technology keynote.

Stay up to date on the latest in business payments by subscribing to our blog! Simply hit the “Subscribe” button above or submit your email address in the form below.

The information in this blog post is for educational purposes only. It is not legal, tax or investment advice. For legal, tax or investment advice, you should consult your own legal counsel, tax, and investment advisers.