How are other municipalities using AI in property management?

Copenhagen, November 12, 2026

Where do you start with AI in municipal property management? And how do you ensure that the results are reliable?

These were some of the questions addressed on November 10, when proprty.ai brought together Rudersdal Municipality, Ishøj Municipality, and NIRAS for a webinar on AI in property management.

The panel consisted of Bryan Morillo Karlqvist, Project Manager in the Property Department at Rudersdal Municipality; Jens Godwin Damgaard, who works with operations and maintenance in Ishøj Municipality; and Michael Ørsted, Project Manager at NIRAS. Rasmus Sørensen from proprty.ai served as moderator.

The conversation provided concrete insights into how municipalities are already using AI, the benefits they are experiencing, and why the technology should be viewed as a supplement to professional judgment rather than a replacement for it.

Start with what is most annoying

Many municipalities are still exploring where AI can create value. However, experience shows that you can often go a long way with the data you already have.

Michael Ørsted from NIRAS recommends starting with three simple questions:

“Is this something that annoys me? Is this something I have to repeat many times? And is this something where I can professionally assess the result of what comes out?”

If the answer is yes, it is a good place to start. Create a simple plan and begin where the frustration is greatest or where the data foundation is already sufficient.

In Rudersdal Municipality, it is primarily the heavy and repetitive tasks that are at the top of the list.

“I love the agents. It frees up time to actually sit and do something meaningful, instead of us spending three months nonstop entering 17 data points for 12,000 rooms,” says Bryan Morillo Karlqvist.

From Excel sheets to answers in five minutes

Previously, Rudersdal’s building inspections ended up as long lists in Excel. The lists showed a greater need for maintenance than there was funding for, and they were rarely used again.

With AI, there has also been a greater focus on the quality and reliability of the data used for decision-making.

“Suddenly, we are talking about how confident we are in this. We never asked that with the Excel sheets. The Excel sheet would just lie there and die quietly,” says Bryan Morillo Karlqvist.

Today, the municipality can, among other things, use proprty.ai to quickly get an overview of long-term maintenance needs.

“Imagine a property manager standing in the office saying: I need this for a meeting this afternoon. What does it cost to maintain this property for the next 30 years? It took me less than five minutes,” says Bryan Morillo Karlqvist.

The maintenance plan also shows where there may be opportunities to coordinate work. For example, if the roof needs replacing one year and the windows the next, the tasks can be planned together so that scaffolding only needs to be set up once.

This also gives management a better overview of the property portfolio. In proprty.ai, the condition of the buildings can be viewed in relation to their energy labels, among other things.

“Our director always asks what our ten worst properties are. Instead of standing there mumbling, you can say: These are the ones,” says Bryan Morillo Karlqvist.

AI as a supporting tool, not a replacement

In Ishøj Municipality, Jens Godwin Damgaard has built the municipality's maintenance program from the ground up, based on his own condition assessments.

When he plans new doors and windows for a building, for example, proprty.ai can simultaneously point out the roof and other building components that he hadn't included in the plan himself.

“It estimates it as a suggestion, and then I have to decide on it. I don't have to use it, but I am forced to take a stance on it,” says Jens Godwin Damgaard.

For him, the value isn't just about saving time, but to a large extent about improving the quality of the decision-making basis.

Among other things, he can see how reliably a building's condition has been assessed. The certainty is higher for the buildings he has inspected himself, and lower for those he has not yet been out to see.

“We can't just set AI to create our maintenance plans. That doesn't work. But AI is a huge supporting tool when we need to assess things,” says Jens Godwin Damgaard.

Both municipalities also use Tily, the AI agent built on top of the domain-specific AI in proprty.ai.

When Ishøj's energy colleagues needed to investigate which roofs were suitable for solar panels, Jens Godwin Damgaard asked Tily to create an overview of the remaining service life of the roofs. The result covered far more buildings than he had managed to assess himself.

A better basis for dialogue with politicians

One of the major challenges in municipal property management is creating an accurate picture of long-term maintenance needs.

Jens Godwin Damgaard has therefore compared his own 30-year maintenance plan with the calculation in proprty.ai. The two calculations are far apart.

“I don't know which figure is correct, but it at least gives me a tool to create a more precise basis for what our needs actually are,” says Jens Godwin Damgaard.

In Rudersdal Municipality, working with AI has also changed the way they talk about maintenance budgets.

“The paradox in municipalities is that we get a pot of money, and then we spend it where it's needed most, we hope. Now we are turning things upside down and saying: With a certain degree of certainty, this is the minimum amount of money we need to spend,” says Bryan Morillo Karlqvist.

Michael Ørsted from NIRAS sees the same potential in being able to visualize the consequences of different priorities before decisions are made.

“We might be more confident in making overarching decisions if we can see what consequences they will have,” says Michael Ørsted.

Colleagues, data, and IT must be involved from the start

Technology is one thing. Getting the organization on board is another.

During the webinar, one of the questions raised was how municipalities can get employees to adopt new tools.

In Rudersdal, it has been a matter of changing employees' habits and the way they work with building data.

“It is a huge step from registering damage to suddenly having to consider the overall condition,” says Bryan Morillo Karlqvist.

In Ishøj, Jens Godwin Damgaard chose a more practical approach. He took the entire construction department to a kindergarten, where employees worked in pairs to perform a condition assessment directly on their phones. Afterwards, they discussed the results together.

“That combination of assistance and attitude helped them understand that there is a tool available that they can actually use in their daily work,” says Jens Godwin Damgaard.

Michael Ørsted also highlighted the importance of incorporating data and IT from the very beginning.

His advice was not to turn data cleansing into a massive cleanup project. Start by gathering the data that already exists and let the discrepancies show you where action is needed.

Involve IT early, limit the solution to the data the task requires, and make sure to keep personal data out of it.

As Michael Ørsted emphasized:

“You should be able to do this, even if your AI isn't working.”

AI in property management starts with the right questions

The webinar showed that there is no single recipe for how municipalities should implement AI in property management.

However, the experiences from Rudersdal and Ishøj point to some common opportunities: less time spent on repetitive tasks, a better overview of building conditions, and a stronger foundation for prioritizing maintenance budgets.

But technology does not do the work alone. Professional judgment, employee experience, and the quality of the data foundation remain crucial.

Therefore, it is not about letting AI make the decisions, but about giving employees better tools to make them.