It’s probably the biggest technological shift since the internet, and for some, its impact could eventually rival the invention of print itself. Yes, we need to talk about AI.

Discussion around AI often falls into two camps: the evangelists who see it as the answer to every business challenge, and the doomers convinced it will eventually take over everything. But what does AI mean for XMPie customers? Where is AI creating genuine value today, and where is the hype running ahead of reality?

To find out, I spoke with Nahum Cohen, XMPie’s SVP of R&D and Customer Outcomes. Nahum oversees the teams that build, evolve and support the platform and is helping shape how AI is being integrated into XMPie’s personalization, web-to-print and omnichannel solutions. I asked him about what XMPie is actually doing with AI today to help customers increase revenue and get more from their investment.

 

Strengthening the Value of XMPie

“The real challenge isn’t writing code. It’s building a platform organizations can depend on to deliver critical customer communications.”

 

Let’s start with the obvious question. Is AI a threat to a company like XMPie?

We see AI as accelerator and an opportunity rather than a threat. It’s a powerful technology that can enhance what we do and help our customers deliver even better results. We’re already using AI internally, incorporating it into our product development, and continually evaluating how it can help customers tackle real communications challenges more effectively.

There’s a lot of talk about AI as if it can do everything, but software is about much more than generating code. We’ve spent more than 20 years solving real business, production and workflow challenges for our customers, and that knowledge is embedded throughout the platform. You can’t recreate that overnight. So we’re taking a practical approach to AI, focusing on where it genuinely adds value and being honest about where it doesn’t. For us, the opportunity isn’t to replace the expertise that’s already in XMPie; it’s to use AI to make that expertise even more valuable for our customers.

 

If AI can build software, why does a platform like XMPie still matter?

Building software and building software platforms are not the same thing.

If somebody wants to create a simple workflow, a dashboard or a lightweight limited application, they can absolutely do it. They could do it before AI, and now the AI tools make this option dramatically easier than it was say two years ago, and we can’t ignore that. But easy to build and ready for production are two very different things. When you’re talking about a production platform, the challenge isn’t just writing code. Running a mission-critical communications platform year after year is something else entirely.

A personalization platform has to be reliable, scalable, secure, and maintainable. It needs to handle edge cases, accommodate changing customer requirements, survive upgrades, and integrate with complex technology stacks. It needs governance, support, and accountability.

Those are not challenges that disappear simply because AI can generate code more quickly than humans and that’s the part that unfortunately, tends to be overlooked. The real challenge is building something organizations can depend on no matter what. That’s where years of engineering investment and experience matters, and why platforms like XMPie offer so much more than the code itself.

 

So, while AI doesn’t remove the need for a robust platform and the expertise behind it, does it make sophisticated capabilities accessible to a much wider range of users than ever before?

Absolutely. We’d be kidding ourselves if we pretended otherwise. So many tasks that previously required specialist developers can now be done much faster with AI – but that’s something to celebrate. It echoes the very reason XMPie was founded: the idea that sophisticated personalization shouldn’t be limited to technical specialists but should be within reach of a much broader audience. We wanted marketers and designers to be able to create highly targeted, data-driven communications without needing to write code or rely on developers for every change. AI pushes that idea even further, which is why we see it as an opportunity, not a threat.

Where I think people sometimes get carried away is assuming that creating something is the same as running it. Of course, AI can help you build a solution remarkably quickly. But creating a workflow or automating an InDesign template isn’t the same as delivering high-volume, production-grade variable data communications. VDP for instance isn’t just about generating output, it’s about optimizing data, workflows and print streams at scale. That’s the result of decades of engineering and real-world experience. Someone still has to secure it, maintain it, support it, integrate it with other systems, and adapt it as requirements change and that’s where the real cost lives. When you’re estimating a project, are you thinking about the first six months or the next ten years?

Our customers aren’t looking to become software companies. They want to communicate more effectively, drive better results for their clients and solve business problems. So rather than asking them to build and maintain their own tools, our job is to bring the benefits of AI into a platform that’s already proven, supported and scalable.

 

Putting AI to Work

“AI doesn’t replace expertise, it amplifies it.”

 

How is AI changing how you build XMPie?

There isn’t a part of the business that hasn’t benefited in some way from AI. We’ve been using it across all departments, and it’s made us noticeably faster and more adaptable.

Our support teams can find information faster to resolve challenges. Our developers and QA teams can validate things more efficiently. We’re able to remove a lot of the friction around mundane processes and focus on innovation instead. It all helps us to deliver a better experience for our customers.

We’re also bringing AI into the product line. In StoreFlow, there’s now an in-app AI Assistant — powered by Chatbase — right inside the Backoffice, helping administrators navigate complex workflows, find answers quickly and get more done with less effort. And in Reports, our new AI Assist feature lets users describe the data they want in plain language and have the platform build the report for them — a first-party example of AI turning intent into a working result on our own platform.

 

Do customers need to wait for XMPie to build more AI into the platform before they can start taking advantage of it?

The assumption that you need to wait for software vendors to package AI into a product before you can use it doesn’t fit with the way XMPie has always worked: if a capability exists elsewhere, you should integrate it rather than wait for it to be built in.

We’ve spent years investing in APIs and integrations because our customers don’t operate in isolation, and neither should their technology. AI is just the latest example.

At our recent Innovation Days, we showcased integrations with services such as OpenAI, including DALL·E for image generation demonstrating how customers can bring AI-generated content and imagery directly into personalized omnichannel communication workflows. These integrations demonstrate something we’ve believed for a long time: customers get the most value when they can connect new technologies into their existing workflows as soon as they’re ready, rather than waiting for them to appear on a product roadmap.

Another example is our Adobe Express integration for StoreFlow, which among other capabilities offers AI-powered creativity in a web-to-print ordering environment. Users can create and personalize print products using dynamic content and intelligent design assistance, without ever leaving the browser. It’s another illustration of our approach: making new technologies always accessible within the workflows our customers are already using.

But I don’t think it’s realistic to expect any one platform to deliver every AI capability. The technology is evolving too quickly. What’s more important is having a platform that can connect these technologies together, so customers can experiment, adopt the tools that work best for them and continue managing, personalizing and delivering their communications. That’s been our approach and it also lowers the cost of experimentation, which is one of the most underrated aspects of AI today.

 

Unlocking More of XMPie

“AI lowers the barrier to experimentation, not just reducing the cost of development but reducing the cost of curiosity.”

 

You mentioned that AI lowers the cost of experimentation. What do you mean by that?

Over the years, we’ve always had customers who wanted to do more with the platform than they were actually doing. The limitation often wasn’t XMPie itself; it was the cost, time and technical effort required to explore an idea. A customer might think, “It would be great if we could build this workflow, connect these systems, or create this custom application,” but once they estimated their development effort, the idea stayed on the wish list.

What’s changed is that AI lowers the barrier to experimentation, not just reducing the cost of development but reducing the cost of curiosity. People can prototype ideas more quickly, explore integrations they might never have attempted before, and test concepts without making a major investment upfront.

 

What role do you see it playing in the actual creation of customer communications?

AI is already proving incredibly useful for content creation. It can generate copy, create images, suggest variations for different audiences and help people move from idea to execution much more quickly. But creating the content is only part of the job.

Once it’s been generated, you still need to make sure it’s accurate, on-brand and compliant. You need to decide who receives which version, through which channel and when. You also need to manage approvals, customer data, business rules and reporting, you need to deploy your variants across thousands of customer journeys.

So again, I don’t see this as a choice between AI and a personalization platform. They solve different problems. AI helps create the content. XMPie helps manage the data, business rules, workflows and delivery that turn content into customer communications. The real opportunity comes from combining the two.

 

Advances in AI cuts both ways. Cybercriminals are using AI to find vulnerabilities and launch attacks more effectively. How are you responding to that?

Absolutely. Cybersecurity has always been an arms race. AI is making attackers more efficient, but it’s also giving defenders better tools to identify risks, detect threats and respond more quickly. Our focus is on staying ahead of the curve.

We use AI to continuously test our applications and APIs, to strengthen security reviews during development, to improve threat detection and to help identify potential vulnerabilities earlier. That’s all layered on top of our existing security program, including SOC 2 Type II compliance, Xerox’s security oversight and independent penetration testing.

We’re finding and fixing potential threats faster than ever, and AI has become a very powerful tool for doing exactly that.

 

A Platform for the AI Era

“AI raises the floor for everyone. But if you’ve spent decades building deep capabilities and solving complex customer problems, it also raises the ceiling too, unlocking even greater possibilities.”

 

So, AI isn’t levelling the playing field at all, it’s reinforcing the value of the platform you’ve already built?

Exactly. AI is validating what we’ve championed since XMPie was founded: sophisticated personalization shouldn’t be limited to technical specialists.

For more than two decades, XMPie has been helping customers create highly targeted, data-driven communications without needing to understand all the technology underneath. AI takes that a step further. It lowers barriers, accelerates experimentation and makes powerful capabilities accessible to more people. That’s a natural fit for XMPie because we’re a platform built to be built on. We’ve always invested in APIs, integrations and openness, giving customers the flexibility to connect new technologies as they emerge.

What AI doesn’t change is the importance of the foundations underneath. Reliability, scalability, security and years of real-world experience still matter. In fact, I’d argue they matter more than ever. XMPie has never tried to do everything itself; we’ve focused on providing a powerful communications platform that customers can build on. AI simply makes it easier for customers to connect systems, automate processes and unlock more value from it, while XMPie continues to handle the security, conformance, production efficiency and scale underneath.

AI makes us faster. It makes our customers more productive, and it allows us to build on more than two decades of investment in the platform. AI raises the floor for everyone, but if you’ve spent decades building deep capabilities and solving complex customer problems, it raises the ceiling too. If anyone should worry about AI, it’s the weaker players, not the platforms with depth.