Building the Future of AI in Payments: A Conversation With Danny Shader & Dave Grannan
Throughout our recent series on AI in payments, we’ve explored how AI is already being used across payment experiences, where we believe agentic AI can take the industry next, and how PayNearMe is approaching that opportunity.
One theme has remained consistent: our clients aren’t asking for AI because they want the latest technology. They want to operate more efficiently, reduce manual work and create better payment experiences for their customers.
AI has the potential to do just that when implemented correctly, but realizing that potential takes more than bolting an agent onto an existing payment system. Agentic AI needs access to the data, business rules, workflows and payment capabilities required to take meaningful action. It also needs to operate reliably within the security, compliance and operational guardrails that come with moving money.
That’s why we believe the future of AI in payments will be shaped not just by who can build the smartest agent, but by who can put that intelligence to work inside the right platform. That brings us to why we have acquired Marr Labs.
PayNearMe has spent 17 years building the payment infrastructure and industry expertise behind PayXMTM. Marr Labs brings deep experience building and operating sophisticated AI agents at scale in regulated industries. Bringing those strengths together gives us an opportunity to turn many of the ideas we’ve explored throughout this series into real-world payment experiences.
To talk more about why the companies are coming together and what the combination could mean for the future of payments, we sat down with PayNearMe Founder and CEO Danny Shader and Marr Labs Co-Founder and CEO Dave Grannan.
Why this acquisition and why now?
Danny: The next major opportunity to reduce the total cost of payment acceptance lies in automating more of the touchpoints across the payment journey. AI creates an obvious opportunity to do that. But doing that successfully requires deep integration between the modules of the payment system and the corresponding AI-powered agents that will drive them. The success of those efforts involves critical but subtle nuances in the integration which in turn require deep understanding of both the technologies and the business processes you’re trying to automate.
PayNearMe brings that domain knowledge and the underlying platform. The Marr Labs team brings deep experience operating sophisticated AI agents at scale. The combination makes a ton of sense, and we’re really excited about it.
Dave:
At Marr Labs, we built and operated AI systems at enterprise scale, including deployments handling more than two million calls a month. What we learned is that agentic AI creates the most value in exactly the kind of work payments are full of: high-volume, high-stakes interactions where accuracy, compliance and customer experience all have to hold at once.
The question for us was where that expertise could have the greatest reach and leverage. PayNearMe has a large client base, a platform those clients already trust and a deep understanding of how money actually moves. That creates a tremendous opportunity.
What did each of you see in the other company that made this the right fit?
Danny: If you only have the technology, I don’t think you can win. And if you only have the clients and client knowledge, you can’t win. You need both.
Having clients isn’t just about commercial relationships. It’s about truly understanding their business problems, and then applying the right technologies to solve them. That’s how we’ve succeeded. Now that we’re expanding into the AI domain, the Marr Labs team brings a depth of experience few companies in the world possess.
Although the Marr team is wicked smart, intelligence alone isn’t enough, nor is a great demo. It is very hard to build agentic systems that work in the real world. Marr Labs has done that.
Dave: For us it wasn’t just the installed base — plenty of companies have clients. What stood out was how PayNearMe thinks. This team obsesses over reliability, compliance and the fine detail of how their clients’ businesses run, and that’s precisely the mindset it takes to operate AI agents in production. When we saw PayXM up close — the business rules, the data, the payment capabilities — it was clear this was the platform our agents were built to plug into.
What are PayNearMe clients really asking for when they say they want AI?
Danny: Our clients are hearing that they need AI, but nobody really wants AI for AI’s sake. What they are really saying is, “I want to make sure my business is as productive as it can be.” They need a partner that can think through the related problems to deliver useful solutions rather than simply handing them another tool.
Trust is a huge part of this. Our clients trust us to move money and help run essential parts of their businesses. We take that responsibility seriously. They come to us with business problems and know that we will find a creative, responsible way to solve them.
Why does agentic technology belong inside the payment platform instead of being bolted on?
Dave: Agents are a new interaction model for software—a shift on the scale of the web or mobile —and leaders in every vertical will bring that capability inside their platforms rather than bolting it on.
In payments, the value comes from pairing natural conversation with everything already embedded in PayXM. The agent isn’t a separate product sitting next to the platform. It‘s the most intuitive way for clients and their customers to use it.
Danny: Exactly. Clients can’t succeed by bolting a generic agent onto a payment system and then trying to stitch everything together themselves. First, it’s not their skillset, but second, the interactions that need to be automated are multi-modal, and some of those modes exist in the payment platform rather than the bolted-on agent. For example, the agent must provide a natural way to access the business rules, workflows and capabilities already built into the platform.
When an agent applies those business rules through voice or another conversational interface, it becomes the interaction layer for all the knowledge underneath. If you put an agent on top of a legacy payment platform, you combine a suboptimal set of capabilities with a loosely coupled agent. That’s kind of the worst of all worlds. Great experiences come from combining the interaction with a platform optimized to solve the real world problems that our clients have explained to us in detail, and that we, by exposing our platform through agents, are uniquely capable of solving
What separates a compelling AI demo from a system that actually works in production?
Danny: Every tech exec knows that sexy demos are easy, because demos can ensure nothing goes off-script. That’s not what the messy real world is like. Our experience has taught us that the difficult work lies in preparing for what happens when things don’t go as planned. Agentic systems are similar. The challenging work is in handling the edge cases, managing response times, testing behavior, monitoring performance and establishing the guardrails that keep the experience reliable…and compliant.
Dave: Much of computer science has historically been deterministic: the code does what you tell it to do. AI systems are probabilistic, so you have to approach them differently.
Our team’s experience with probabilistic systems goes back to speech recognition. We learned how to make those systems perform reliably in production. With agentic AI, the goal is to combine probabilistic intelligence with deterministic controls, giving people a natural, flexible experience while still producing dependable business outcomes.
Why is that especially difficult—and especially important—in payments?
Danny: “Deterministic results from a probabilistic world” is a good way to describe success.. That is hard, especially in payments.
We are in the business of completing payments and moving money, so we hold ourselves to a very high quality bar. We have earned our clients’ trust through quality and reliability, and we must remain unwavering about those standards as we apply AI. Guardrails around security, compliance and observability, as well as deterministic controls, are critical.
Dave: There is a real tension between guardrails and user experience. You can make a system very rigid, which makes it easier to control, but people don’t naturally speak in a rigid order. They interrupt, change topics, answer a different question or circle back. A useful agent has to follow that conversation naturally without losing context or stepping outside the rules.
That creates a three-dimensional challenge: compliance, conversational flexibility and performance. You need to excel at all three. Marr Labs was built for regulated industries from the beginning because we wanted to take on the hard use cases, where every error is costly and quality really matters. That makes PayNearMe a natural home for our technology and team.
Where does agentic AI fit within Payment Experience Management?
Danny: Payment Experience Management is all about understanding and optimizing the entire payment experience and continually finding ways to make it better. Agentic AI is an enabling technology for Payment Experience Management that gives us an entirely new set of tools for managing and improving every step of the payment journey.
What makes agentic AI particularly powerful is its ability to move beyond responding and explaining to taking meaningful action. By integrating that intelligence into PayXM, we can address complexity across customer self-service, support interactions and back-office operations. That in turn can help businesses accelerate payment collection and reduce the total cost of payment acceptance.
Once agentic capabilities are embedded in the platform, what becomes possible for clients?
Danny: One of the most exciting opportunities is the data advantage. The Marr Labs team has learned from millions of real-world interactions. PayNearMe has deep, industry-specific knowledge of how and when people engage and make payments. Combining insights from those experiences helps us deliver agentic systems that drive the business metrics our clients care about most.
As we deploy these capabilities across the platform, performance data can help us evaluate and refine them within the appropriate privacy, security and data-governance boundaries. Over time, agentic AI can help us perform existing work much more effectively while creating opportunities to solve additional payment problems.
How will Marr Labs influence the way PayNearMe builds, learns and operates?
Dave: Marr Labs has been an AI-native company from the beginning. That has shaped not only what we build but also how we work: rapid iteration, machine-learning pipelines, evaluation, observability and continuous learning from data.
We’re bringing that experience to PayNearMe, and we’re equally excited to learn from a team that has spent years building a trusted platform for a complex, regulated market. The best version of this combination is one where each team makes the other better — and that’s already happening.
Danny: The Marr Labs team brings not only product expertise but also a different way of operating. Its AI-first practices can help propel us forward, not only in the agentic products we deliver but also in how we develop software, evaluate results and continuously improve.
What should clients and the market take away from this acquisition?
Danny: Our success, both to date and in the future, derives from a magical combination: deeply, deeply understanding our clients’ problems and applying the right technology to solve them.
This acquisition is the next step in that journey. We already understand the payment problems our clients face. Now we are adding a remarkable team and a powerful new set of tools to solve those problems in ways that simply weren’t possible before.
Dave: Clients don’t actually want AI — they want payments collected faster, costs coming down and customers who end the call satisfied. That’s the standard we hold ourselves to. We’ve spent years learning what it takes to make agentic AI deliver real results in production, and PayNearMe gives that expertise the platform and reach to matter. Together, we’re going to make agentic payment experiences something clients can simply count on.
