The real frontier smells of steel, of earth, of building material.

AI is the key. Traditional industries are the lock.

The aging wave is an existential crisis

A quiet crisis is unfolding in nearly every traditional industry. In the second quarter of 2026, 461,000 of the 1.66 million people working in Dutch manufacturing, construction and transport were 55 or older. More than one in four.

They take something with them that no ERP system ever stored. Decades of judgment. The planner who knows how to restart a supply chain when everything fails at once. That knowledge is in no manual. It is in people, and those people are leaving.

Silicon Valley looks at itself. Meanwhile fewer than a third of Dutch industrial firms with ten or more people use any form of AI. The one tool that could keep that knowledge is barely in the building.

A company that solves this does not sell software. It sells continuity. Nobody beats continuity on price.

Their IT is a time bomb. You are the bomb-disposal expert

Walk into any traditional company. Not the lobby. The operation. The factory floor, the building site, the warehouse.

Three in four Dutch industrial firms with ten or more people already run an ERP system. Often one from 2001 that is “actually quite stable”. Excel files are the nervous system of million-euro operations. The system is not what’s missing. The seams are. Quoting a custom machine. The engineering knowledge behind last year’s design. The service data from machines in the field. All of it lives in Excel, in inboxes, and in the head of someone who retires in three years.

For these companies IT was always a cost center, never a weapon. What’s left is layer on layer of patched legacy, workarounds nobody wrote down, and knowledge stored nowhere.

The most valuable AI companies of the next decade won’t build models. They will build what finally makes the industrial world intelligent.

AI automates judgment. Industry runs on judgment

Ten years ago, automation meant repetitive tasks. AI now automates judgment, and traditional industry runs on it.

The inspector who spots a defect on the line. The planner who rebuilds the schedule when a machine breaks down. The site supervisor who sees a safety risk in a photo. None of this was out of reach because it was complex. It was out of reach because it was unstructured. AI removes that barrier, and it can capture some of what the retiring generation knows before it walks out of the door.

This is not a rescue mission

None of this is charity. It is where the money is.

In 2025, agriculture, industry and energy, construction, transport, and health and social care added €377 billion to the Dutch economy. That is more than a third of everything it produced. Not the app stores. Not the ad networks. The companies that pour concrete, distribute medicine and move containers from A to B. In construction alone, one percent of the value the sector adds is worth more than half a billion euros a year. Agricultural exports alone came to €137.5 billion.

These are not leftovers of an industrial age. They are cash-generating mastodons with structural budgets, long contracts and a modernization problem they cannot solve themselves.

The customer has money. The customer has urgency. The queue of serious contenders is remarkably short.

Incumbents want to change. Their immune system won’t let them

Talk to anyone who runs a traditional company and you hear the same thing. They know they have to change. Their organization is built to resist it.

Not because the people are slow. A company that runs a factory or a fleet has to run it safely every single day, and that makes it careful by design. The CFO can’t sign an investment that hurts this quarter’s margin. The operations manager can’t gamble a running line on an unproven pilot. So they open an innovation lab and go to Davos. The lab rarely survives the trip back to the operation.

They create the demand. They can’t build the solution alone. That is your opening.

Geopolitics works for you. Use it

After COVID, the chip war and the energy crisis, Western governments rediscovered something they had forgotten for decades. Industrial sovereignty is strategic.

The European Chips Act. The American CHIPS Act. Dutch public money for manufacturing and the energy transition. Billions are now set aside for exactly the sectors where we build. Political tailwind, fresh capital and pressure to modernize, all in the same sectors, all at once.

When anyone can build, building stops being the edge

AI made building a product cheap for everyone. After late 2022, AI coding tools pushed more people to launch a first venture. Those newcomers were also less likely to survive.

In traditional industry, the edge is trust. In Europe, 92.4% of deals go to a vendor that was on the buyer’s shortlist from day one. Buyers ignore unsolicited outreach on purpose. A better product sent cold does not get the meeting. A product from someone the buyer already knows does.

AI levelled the build. It made the relationship the scarce part.

No outsider can buy what the insider knows

Every machine builder has an engineer who knows exactly why quoting takes three weeks. Every distributor has a planner who rebuilds the same broken schedule every Monday. They have looked at the problem for years. They know who would buy the fix, and what it would have to do on day one.

A founder from outside needs a year on the floor to learn what the insider already knows. The insider also has the thing that matters most. Inside their company, the buyer already trusts them.

Prove it before anyone bets the house

The startup world has one entry rule. Quit first. For a 40-year-old engineer with a mortgage and a family, that is not courage. It is a bet with the household’s money, and many sensibly decline it. The Dutch entrepreneurship monitor names exactly this group, older and highly educated, as the country’s untapped potential. They have “a more attractive and secure alternative in paid employment”.

The evidence says their caution is right. A study of American workers tracked people who started a business while still employed. They gave it up about a third less often than people who quit first. Proof first is not the timid route. It is the one that lasts.

The people who know where industry breaks should not have to gamble their family to fix it.

Regulation is not your enemy. It is your moat

A naive founder sees regulation as an obstacle. A smart founder sees a moat.

Healthcare has the GDPR, the WGBO and medical device law. Finance has Basel IV and MiFID II. Food has HACCP, EU traceability rules and the rules of every country it ships to. A horizontal AI tool from Big Tech cannot find its way through that.

A startup that breathes the rules, that builds compliance into the product and the workflow, builds a defense capital alone cannot buy. Once a customer runs on your compliance, the cost of switching is enormous.

AI that already knows a sector’s rules is one of the most underbuilt products there is.

Vertical wins. Always. From now on

For ten years the tech world believed in the horizontal platform. AI ends that belief.

A generic tool that does everything a little loses to a sector system that does one thing perfectly. Not by a margin. Fundamentally. The model is not the edge, because every frontier model is available to everyone. Train a system on thousands of construction drawings and contract disputes. Build it into the way a site actually runs. On that site, it beats a generic chatbot every time.

The sector data is the moat. The domain understanding is the product. The relationship is the defense. Vertical companies build the one thing the largest tech companies in the world cannot match. Depth.

Software was built for people. Agents don’t need it

Every software product of the past thirty years assumes a person has to operate it. Hence the onboarding flows, the dashboards and the training. An agent needs none of that. It needs capabilities it can rely on, a data model, and the sector knowledge to combine them. To an agent, the interface is a steering wheel in a car that drives itself.

Some companies have already reorganized around this. Block builds capabilities like payments and lending with no interface of their own. An intelligence layer combines them for one customer at one moment, and no permanent layer of middle management sits in between. Traditional industry is full of processes that can become capabilities, like quoting, scheduling, procurement and inspection. Someone who knows the industry from inside can see which ones. Someone building a pretty dashboard from outside cannot.

What feeds that intelligence is a world model. It holds what the company knows about its sector: the transaction data, the process data, and the judgment of the people who do the work. It gets richer every day the system runs, and nobody outside the sector can copy it. The capabilities are the building blocks. The world model is the moat.

Whoever still builds a dashboard for traditional industry is solving the wrong problem. It needs an intelligence that does the work.

We run the studio this way

This is not theory. We run the studio on a world model. Every weekday an AI agent reads our meetings, mail and documents, updates what we know, and flags what needs a person. We would rather read than meet. People stay at the edge, where judgment and relationships live.

It is how we think every new company in industry should be built.

The closed door

Look at it from both sides. The insider sees the problem but cannot risk the job to try. The founder from outside has built the product, but the buyer has never heard of them. One has the insight and no way out. The other has a product and no way in. Industry stays as it is, and the fix gets built by someone the buyer already trusts, years later, or never. A fund can back a founder who has already left. We work with the ones who haven’t, and bring the buyer with us.

Our half

Industry buys from people it knows. Since 2012, a lot of it has come to know us. OWOW has designed and built digital products for industrial clients, some of them for years. That trust is what we put behind every venture, so we hold ourselves to one rule.

We only invest when we can find your first customers ourselves.

14

years building for industry

500+

active clients

€2.65M

raised for ventures

Before the studio opened, Ver.iD’s founders came to us with no product or revenue, and we built the first full-scale platform with them. With Meridian’s founders, we designed and built its apps. Today we are building Synthgen, synthetic data for training AI in robots and machines, from Eindhoven.

Our bet, and how we will know

We have not proven this yet. Our case rests on public data and fourteen years of client work, not on an industrial exit. We bet that companies started by insiders, with buyers who already trust them, find customers faster and last longer than companies started cold. Synthgen and the ventures after it are the test.

If they don’t, we were wrong about something on this page, and the next version will say what.

You know where your industry breaks. We co-found the company that fixes it.