Many organisations are experimenting with Gen AI to accelerate processes and support employees more intelligently. Yet it often remains at isolated pilots: the technology works, the possibilities are there, but real impact is lacking. Not due to a lack of tools, but because a clear organisational form is missing. Without clear ownership and fixed frameworks, Gen AI gets bogged down in isolated initiatives. Therefore, the key to success rarely lies in even better technology, but in the way you organise Gen AI and focus on adoption. Julie van Nierop, Head Technology Innovation at Pon and responsible for the construction and implementation of Gen AI projects together with NXTminds, explains how they tackle it.
Many companies start with great enthusiasm. But after the initial successes, noise arises: who is the owner? How do you scale up? And how do you ensure that AI does not remain an experiment of a few enthusiasts?
AI starts with the problem, not the solution
According to Julie, that is no coincidence. “Gen AI is a tool,” she says. “Not an end in itself. The real question is: what problem do you want to solve? Then you look at the process and only then do you see how technology can help.” At Pon, where Julie has been working for three years, that order is sacred. Pon is a family multinational with 11,300 employees in 32 countries on six continents. From the Datalab, Julie works on AI and Gen AI solutions that are not only smart but also focus on the adoption of these solutions
“Too often, organisations start with the tool,” she says. “Then the question becomes: can we stick AI on this? While you should actually first look at what you want to solve and the customer need.
A common example is customer service. “Organisations receive large volumes of recurring questions, while employees mainly spend their time on repetitive work,” Julie explains. “This comes at the expense of speed, quality, and attention to more complex customer questions.”
“In this situation, you first look for a solution,” she continues. “This can start with automation, for example with a chatbot. Then this can be supplemented by automating other customer touchpoints, such as email and voice. But every step requires different choices; in process, technology, and organisation.”
Programme or Centre of Excellence?
Once organisations take AI more seriously, they often choose a programme approach. With a fixed budget, clear goals, and an end date. “That makes sense,” says Julie. “A programme helps to create focus and speed.”
But Gen AI is not a temporary subject. That is why Pon consciously chooses a combination. “At Pon, we have a Centre of Excellence where experts are continuously working on making customer needs, data, and Gen AI transparent. That knowledge remains available. At the same time, we run programmes where we temporarily give certain subjects extra attention in the organisation.”
Within the Centre of Excellence, different teams work together: data experts, Gen AI specialists who build bots and platforms, and people who think from the customer’s perspective. Everything comes together on one platform, set up according to fixed security and privacy standards.
“The programme that results from this is temporary,” Julie explains. “Six months, fixed objectives. We accelerate use cases from the business, train people, and conduct a scan to then give teams tools to scale up further.”
From pilot to adoption
That distinction, between building something and facilitating adoption, is crucial. “Many AI initiatives get stuck in pilots,” says Julie. “Not because the technology fails, but because adoption is not organised.”
At Pon, AI adoption is consciously organised in two ways. “Firstly, we select carefully: there must be a business case behind it, and we always check whether a solution is reusable for other Pon companies and in line with Security and Privacy standards. Additionally, change management plays a major role. “We train people at different levels and have built various curricula.”
Why it gets stuck if no one is the owner
AI changes how people work. “The biggest gain is not in the tool, but in behaviour,” says Julie. “People need to adjust their way of working. That takes time and attention, which you have to organise.” Companies that leave AI entirely to the business often get stuck. “People don’t know where to start or get overwhelmed. And without clear frameworks, risks arise, for example around privacy and security.”
Moreover, behavioural change requires exemplary behaviour. “Leaders of the organisation must show that this is important. If it is not explicitly expected, it fades away.”
End-to-end support in AI adoption
In the phase where organisations want to scale AI from isolated initiatives to structural adoption, Pon regularly uses the expertise of NXTminds. Julie: “NXTminds understands that this is not just a technology issue. It’s about people. About profiles who want to learn, who breathe innovation, and help organisations change.”
NXTminds supports end-to-end: from finding the right AI profiles to deploying them in programmes and teams. “That temporary expertise is indispensable,” says Julie. “You can’t always solve everything internally. Especially not in a market that moves so quickly.”
Why organising now pays off
According to Julie, now is the time to think about the organisational form of Gen AI. “It affects everyone. Jobs are changing. If you don’t organise it well now, your competition will, and they will be ahead.”