In the guest blog series of Bits and Bricks Rasmus Juul-Nyholm shares his view what AI will do to the real estate. Rasmus Juul-Nyholm is one of the founders of Proptech Denmark, and an experienced Managing Director and entrepreneur with a strong track record in real estate, IT and finance. Having grown and sold Denmarks's largest service provider Cobblestone to become part of the Nordic INNA Group, he is nowadays engaged with Home.Earth.
------------
I want to share two images that describe how this moment feels to me personally. The first image is of a calm beach where the water has pulled back. People are walking around, slightly confused, trying to understand what is happening. Out at sea, the wave is building. The second image is a dragon flying into our industry. You will either have to tame it and ride its back, to new promising destinations, or it might turn on you and burn down your business. These are not forecasts. They are simply descriptions of the sensation I have right now, standing inside a technological shift unlike anything I have seen being at the centre of digital innovation for 25 years.
We have never seen a technology develop and commercialise at this pace. AI providers are releasing new capabilities on a weekly cycle. We are not counting product generations in years; we are counting weeks. The room of opportunities and threats is expanding so fast that it is difficult to see the horizon. This is not about any single AI capability. It is about the rate of change itself, and what it demands of organisations built for slow cycles.
The real estate industry has long been constrained by a shortage of qualified people. Access to good staff is one of the most persistent inhibitors of growth and a major driver of cost. A new AI / human hybrid workforce changes this equation. Within a short horizon, 10% to 30% of a typical real estate organisation’s operational capacity will be agent-based, loosening the binding constraint and allowing organisations to scale without scaling headcount proportionally. The IT department transforms from a staff function, supporting humans with IT, into something closer to the HR department of an agent workforce: building, training, managing, and retiring agents. This is not a staffing question. It is an organisational design challenge that most real estate companies have not yet begun to address.
Three previously expensive activities have become radically cheaper: analysis, specification, and prototyping. In real estate, the most expensive and scarce human resources are the analysts. Investment analysts and strong finance controllers command the highest wages and are, to some extent, the heroes of the industry. Yet their core abilities are precisely where AI is strongest. With generative and agentic AI, research at an impressive level is readily available. Specifying a new solution used to be complex and expensive. Now it is easier to simply demonstrate it, experiment, and decide whether to convert it into a formal project. The governing principle: demo, don’t memo.
Real estate operates on long cycles: a year in finance, three to five years in development, ten to thirty years in asset holding. AI compresses the observe-analyse-decide-execute loop into weeks and months. The challenge is not speed itself but the mismatch between what becomes possible and how organisations are structured. Management routines, governance, and decision authority all assume annual rhythms. Adapting these is a deeper challenge than adopting any single tool.
AI is reshaping both sides of the landlord-tenant relationship simultaneously. On the landlord side, technology has for years been automating tenant service. On the tenant side, the same tools now professionalise every response and negotiation. The result is not a one-sided efficiency gain but an escalating dynamic where both parties upgrade in parallel. The advantage does not come from having AI. It comes from having better data, better integration, and better strategic deployment.
The real estate industry has never been deeply data-driven beyond financial reporting. Dashboards and BI solutions have become available, but extracting insights and acting on them remains a challenge. AI agents on top of existing data create a proactive, always-present interface that can surface patterns, flag anomalies, and trigger actions without waiting for someone to open a dashboard. This may be what finally makes the industry genuinely data-driven.
In the first article in this series, The Strategic Space of Proptech, I proposed a model mapping where technology categories naturally fit based on two dimensions: degree of vertical integration and asset hold horizon. AI pushes to this equation. Along the integration axis, it lowers the threshold. Capabilities that previously required broad and larger in-house teams become accessible to narrower, leaner players. Along the hold horizon axis, AI compresses time-to-value. If decision cycles move from annual to weekly and prototyping becomes cheap, shorter-hold players can extract value from technologies that previously only justified themselves over longer hold periods. The net effect is that AI pulls the strategic space inward, making more of the proptech landscape accessible to more players. In the second article, Why Proptech Moves Slowly, I described how fragmentation, short cycles, and small organisations create structural barriers to technology adoption. AI is uniquely positioned to weaken precisely those barriers. Strategic fit still matters. But the accessible territory on the map is expanding, and players who recognise this shift early will have a meaningful advantage.
