At Innovation Day: AI Made Real 2026, held in Amsterdam on 3 February, Jeroen Manten, Director of IT Commerce & Channels at PostNL, shared how the organization is rethinking customer interaction as expectations evolve. As we all know, these expectations are shaped by customers, but soon they will also be shaped by AI agents acting on their behalf.
Rather than presenting a single AI solution, the talk focused on a broader shift in how organizations need to respond to intent, context and anticipate outcomes. Together with EPAM and Empathy Lab, PostNL is working toward what they call the Intelligent Front Door. It is not a product or platform, but a single orchestration point where customers interact with PostNL.
The real shift is not AI, it’s expectations
AI dominates conversations across industries. New models, faster tooling and smarter assistants appear almost daily. But as Jeroen made clear, focusing on technology alone misses the real change that is already underway.
Customers no longer want to move through multiple systems, channels and interfaces to get something done. They increasingly expect organizations to act on their behalf, with full awareness of context, history and what they’re trying to achieve.
What has changed is that this expectation no longer comes only from people. AI agents are starting to interact with organizations on behalf of customers, and they have no patience for delays or vagueness. An AI agent will not wait for processes to complete, repeat the same request across channels, or accept delays as part of the experience. If an outcome cannot be achieved immediately, it will reroute to another provider that can deliver it faster.
Why omnichannel is no longer enough
Most organizations have invested heavily in omnichannel experiences. Websites, apps, and contact centers continue to improve, often significantly. The challenge lies between those channels.
Even strong omnichannel setups still require customers to carry their own context, repeat their intent, and manage transitions from one system to another. People have tolerated that kind of hassle for a while, even if it’s been annoying. But AI agents just won’t put up with it.
“Even if you do omnichannel extremely well, it won’t be enough anymore. AI agents won’t wait, won’t repeat themselves, and won’t accept friction.”
Jeroen Manten, Director Digital IT at PostNL
In an environment where work is increasingly delegated to autonomous systems, orchestration matters more than access points. Recognizing intent, maintaining context, and delivering outcomes across the organization becomes foundational rather than exceptional.
For PostNL, this realization led to a change in perspective. The issue was no longer a channel problem. It was an orchestration problem.
One entry point, many intelligent routes
Instead of asking how each channel could be optimized, PostNL asked a different question: What if there was a single entry point, and intelligence determined what happened next? Inspired by that thinking, the Intelligent Front Door was born.
From the outside, it is one entrance. Behind it sits a smart structure that routes requests across domains, systems, AI agents and – when needed – human employees.
The purpose is not automation for its own sake. The goal is to remove friction for customers, for AI agents acting on their behalf, and for employees inside the organization.
“The Intelligent Front Door isn’t something you can buy. It’s not a platform, it’s a vision for how the organization works.”
Jeroen Manten, Director Digital IT at PostNL
Intelligent Front Door is an operating concept and a way of organizing capabilities, responsibilities, and experience design. This is where the collaboration between PostNL, EPAM and Empathy Lab is essential, aligning technology, architecture, and human-centered design around a shared direction.
Building foundations before scaling intelligence
Building foundations before scaling intelligence
PostNL has taken a phased approach rather than attempting to scale AI across the organization all at once.
In 2025, the focus was on building initial agents and prompt-based capabilities. In 2026, the emphasis shifts to orchestration within specific domains and to creating an internal registry of AI agents across the company. That visibility is not about central control, but about reuse, coherence and readiness.
The larger step comes in 2027 and 2028, when orchestration across domains and business units becomes possible on a broader scale.
This sequencing brings us to an important insight: without strong foundations, scaling intelligence only increases complexity. The Intelligent Front Door depends on clarity first, not acceleration.
AI made real through early building blocks
While the vision extends into the coming years, PostNL is already translating these ideas into practice. Several AI-powered capabilities are live in production, and one of them was available as a demo during Innovation Day.
Examples include prompt-driven chatbots that move away from rigid flow-based logic, conversational onboarding for new customers, and a shipment agent that allows parcel pre-announcement and label creation through dialogue instead of traditional interfaces.
These solutions are not positioned as the Intelligent Front Door itself. They are early building blocks that demonstrate how intent, context and history can reduce friction today, while preparing the organization for deeper orchestration tomorrow.
Responding as one
The talk closed with a question that resonates far beyond PostNL.
“The question isn’t whether customers and AI agents will interact with us through AI. The real question is whether our organization is ready to answer as one?”
Jeroen Manten, Director Digital IT at PostNL
For PostNL, the Intelligent Front Door provides a clear direction. For others, the answer may take a different shape. What is becoming universal is the need to move beyond channel optimization toward true orchestration, grounded in experience, clarity, and intent.
Curious how this applies to your organization?
At Empathy Lab, we work with teams to explore what orchestration, AI readiness, and experience design mean in practice, across technology, language, and organizational structures.
If these questions are part of your current challenges, we would love to continue the conversation.
Contributor in this article
