AI Unveiled - GainWorks

By Paul Hoffman, founder & AI director of GainWorks

Twelve years in applied AI, six of them as AI officer at the Royal Netherlands Air Force

June 2026 · 5 min read

The core

Everyone is adding AI.
Almost no one redesigns.

Most organizations bolt AI onto an existing process. That process was shaped around people over many years, and adding AI to it changes nothing fundamental. Everything stays as it was, only slightly faster.

Starting over means asking a different question. Not: how do I make my current process faster? But: what would this process look like if I designed it again today, with AI as the starting point?

That difference looks small. It is not. It moves the boundary of what is possible, in ways the old approach never reaches.

The question is not how you add AI to your work, but what your work looks like when you design it around AI.

The difference no one names

Software in a process is static.
AI is malleable.

Software plays a part in almost every process. The only question is: does the process bend to the software, or the software to the process?

Buy software and you arrange the process around the package. Have software built for you and it is slow and costly, and afterwards you are locked in: every change means building again, so everything stays as it is for as long as possible. Either way the process serves the software, and in a fast moving market that is the wrong place to be stuck.

With AI that relationship reverses. You design the process the way it ought to work, together with the people who are responsible for it. No separate training program afterwards, because they shape the process themselves; and adjusting it is always possible, because AI can be molded where a package is fixed. Now the software serves the process, instead of the other way around.

From a process that bends to the software, to a process you design yourself and in which AI supports every step.

That reversal is not a detail. It changes what a process can deliver.

What you gain

Three things that become possible

Once you design the process around AI, more than the speed shifts. Three things appear that did not exist in the old approach.

1. Digital steps without a ceiling

AI scales without limit. In the “old world” a simple rule applies: if an organization wants to grow, it has to hire people. In the “new world” that rule still holds, as long as the work takes place in the physical world. But the moment a step takes place in the digital world, the rule falls away. To an AI it makes no difference whether a step is taken once or 10,000 times. For a client we had an AI application built that tracks the competition. Where an employee used to look at prices and campaigns now and then, the application does it every day, as often as needed, and proposes on its own what the client can set against it.

2. Knowledge that stays in house

An AI does not leave your organization. You train it the way you train an employee, and it never walks out. You can correct it, and part of that improvement happens by itself: as the technology gets better, the way of working you have built up is carried along with it. It is the employee you never have to persuade to stay, in whom you can invest and who pays off year after year.

People come and go. What an AI once knows and can do stays in the organization.

What you gain

3. The process that turns itself into data

AI handles data well, and it records what it does along the way. Every report that is written, every step that is taken, every question that is answered leaves a trail of documentation and data behind. That trail is valuable in itself.

Take a digital employee that writes reports. Alongside the reports themselves a picture builds up: how many are produced, on which subjects, which questions keep coming back. Or one that writes up conversations. Every write-up yields knowledge: recurring themes, lessons, patterns.

The work delivers not only the result, but also insight into the work itself.

The transition

Why it is hard,
and how you do it anyway

For many organizations this is a difficult transition, and that is no reproach. Most people have worked in the existing process for years. Trying to rebuild that team and that way of working in one go is where it often stalls.

There is a calmer route. Do not start with everyone. Find the people who do want to make the change, and make them the owners of the new process. Let the new and the old exist side by side for a while.

As the new process works better and grows, it starts to act like gravity. A way of working that produces results and gives room draws people in by itself, while the old one slowly dies out. You do not have to force anyone. You only have to make the new one good enough.

What it delivers

What this means for your organization

Ask the question out loud.

What if you could deliver services at a scale and speed that is no longer bound by the number of hands?

What if the knowledge and experience stayed in your organization, even when people leave?

What if your work recorded what it does along the way, and made itself better as it went?

These are not promises about the technology. They are outcomes of a design. And they stack: scale, resilience, and a lead in knowledge that only grows with time.

That is the difference between using AI and building an organization that rests on AI.

Further

Where you can begin

The step is not to do everything differently at once. The step is to choose one process and ask the question: what would this look like if we designed it again today, around AI?

That is exactly what we help with. We map where the largest gain sits in your organization, build the business case that goes with it, and guide the design through to execution. Independent, and with people as the starting point.

The future of work will not be decided by who buys the most AI, but by who anchors AI in the organization.