Amsterdam hosted inspiration and the operating manuals to match. Innovation Day - AI Made Real 2026 brought together executives who have been in the trenches, scaling intelligence from experiments to enterprise muscle. Inspired by the four keynotes, this article explores business value, orchestration, enterprise engineering and the evolving relationship between people and intelligent systems.
Setting the tone: elevate AI to core infrastructure status
During his opening keynote, Alex van Gestel (VP, Head of Benelux and Consumer & Services EMEA) traced how top performers graduate from scattered proof-of-concept experiments to MVPs, then to hybrid solutions pulsing through entire business units. AI is treated like core infrastructure, governed, measured, and refined on repeat.
Alex also highlighted a persistent value gap. Many GenAI programs still aren’t paying off. Teams that do succeed weave AI into daily workflows, track outcomes closely and build the structures needed for responsible scaling. He illustrated this with examples across the EPAM portfolio, including personalization, predictive maintenance, intelligent automation and capital optimization. He encouraged leaders to ask how their organizations would operate if AI were integrated into their identity rather than treated as an add‑on.
PostNL: One intelligent doorway for every customer signal
Jeroen Manten (Director of IT Commerce & Channels, PostNL) explained that expectations are changing rapidly. People no longer want to repeat their intent across channels. They expect organizations to understand context and deliver outcomes without unnecessary steps. AI agents will expect this even more, because they operate without patience for delays or inconsistencies.
PostNL is addressing this through the Intelligent Front Door, created with EPAM and Empathy Lab. This approach offers a single entry point where intent is recognized, context is maintained and requests are routed across systems, domains, AI agents or human teams. The goal is to reduce friction for customers, AI agents acting on their behalf and employees handling internal processes. Early use cases are already live, such as prompt‑driven chatbots, conversational onboarding and a shipment agent that supports parcel pre‑announcement through dialogue. The longer roadmap moves from individual domain orchestration in 2026 to broader orchestration across business units in 2027 and 2028.
This perspective aligns strongly with Alex’s framing: AI becomes meaningful when it influences how an organization understands and responds to intent, rather than when it is used only to improve isolated touchpoints.
Albert Heijn: a governed GenAI platform that doubles delivery speed
Akshay Jadhav and Jorrit van de Geer outlined how Albert Heijn advanced its GenAI capabilities inside one of Europe’s most complex retail environments. What began with internal hackathons in 2023 grew into AI Labs, a secure platform that exposes leading models through governed APIs. The platform offers cost controls, guidelines, templates and hands‑on support so teams can experiment while staying compliant. Dozens of use cases now run in production and the platform handles billions of tokens per month.
The platform has enabled valuable improvements. Steijn, the personal food coach, helps people make informed decisions about meals and reduces food waste. A conversational employee assistant simplifies daily tasks for store teams. Developers benefit from Langflow, which allows them to design AI agents in a controlled environment.
A defining part of their approach is structured AI‑assisted engineering. Engineers provide precise instructions to AI agents, review generated components and refine prompts to align with internal standards. This loop accelerates delivery and improves clarity in architecture and implementation. The team now works twice as fast and continues to build momentum.
Albert Heijn’s story shows the impact of secure foundations, clear processes and a culture that sees AI as an amplifier of expertise.
Neurotechnology: The next interface between brain and machine
In the closing keynote, prof. Pim Haselager explored how neurotechnology is bringing AI and brain science closer together. He introduced the audience to brain decoding, where neural networks identify correlations between brain activity and perception, imagery or intention. Research already shows that models trained on visual cortex patterns can infer what a person is seeing, and similar approaches extend to imagined or dreamed content.
Pim also described brain writing: influencing neural activity through stimulation. Clinical examples include deep brain stimulation for Parkinson’s treatment. Newer non‑invasive technologies, including focused ultrasound and near‑infrared devices, introduce possibilities for affecting attention, creativity and emotional state without surgery.
Wearable EEG devices are becoming more practical, which may introduce new ways for people to interact with intelligent systems. Neural signals could complement behavioral input, helping experiences adapt more naturally to intent and cognitive patterns. Pim emphasized the importance of preparing for neurorights, because brain data carries unique sensitivity. Clear rules will be needed around storage, protection and governance as these technologies become more accessible.
This perspective adds a long‑term lens to the event: as interfaces evolve, the relationship between intelligence and cognition will require thoughtful design and responsible decision‑making.
AI becomes real when strategy, engineering & ethics converge
The four talks connected into a broad view of AI Made Real.
- Alex outlined how organizations unlock value through maturity, governance and measurable outcomes.
- PostNL demonstrated how intent recognition and orchestration reduce friction across the entire experience.
- Albert Heijn showed how platform governance, engineering discipline and hands‑on adoption practices accelerate delivery.
- Pim highlighted a future where neurotechnology may influence how people interact with intelligent systems, raising new ethical considerations.
Together, these perspectives present AI as a capability that touches strategy, customer interaction, engineering and emerging interfaces.
Your next four moves to operationalize AI
You can begin to create business‑wide value by taking practical, focused steps.
1. Identify value hotspots where AI can create the most impact. Examples include demand forecasting, dynamic pricing, fraud detection, supply chain optimization or workforce productivity.
2. Select one high‑ROI use case and treat it as a proving ground. Focus on outcomes such as reduced churn, faster revenue cycles or better operational efficiency.
3. Map the data, systems and teams involved. Define how AI will complement human decision‑making and integrate into existing workflows.
4. Pilot the use case, measure results and scale based on evidence. Strong initial outcomes build momentum for broader adoption.
If your organization wants to strengthen orchestration, AI readiness and experience design, Empathy Lab works with teams to connect strategy and execution so AI becomes a dependable part of how the business operates.
Contributor in this article
