Real Estate: Data Silos Without Connections – Where AI Could Help

TVS News AI, editorially supervised, presents Real Estate: Data Silos Without Connections – Where AI Could Help.

“The real estate industry generates enormous volumes of data but generally uses them in isolation.” For Steffen Sebastian, a professor of real estate finance at the International Real Estate Business School (Irebs), this is the industry’s central weakness. The main problem, he says, is not the lack of data. What matters is that the existing information is not connected.

This brings a fairly concrete question to the fore: How can data from the different phases of a real estate project come together in a way that enables AI to extract more from it than it can from isolated collections? Sebastian sees the greatest untapped lever at precisely this interface.

The Bottleneck Lies Between Four Phases

The approach

Sebastian identifies four areas that, in his assessment, need to be more closely linked: planning, financing, construction and operations. Each of these stages produces information relevant to the real estate industry. If the data is handled separately, the connections between the phases remain unused.

This is Sebastian’s core thesis. He is not describing a missing tool as the sole bottleneck, but rather a gap in the workflow. His formulation is strikingly matter-of-fact: The bottleneck is less about technology than about missing connections. If the connection succeeds, construction will “perhaps sometimes also become easier eventually.”

This final sentence makes clear where expectations lie. Sebastian is not promising a ready-made solution or an immediate increase in productivity. He is describing a direction: Connect data and make that connection usable for AI.

Why this is more than a digitalization issue

When each phase is approached in isolation, it remains separate from the others. The real estate industry already has enormous volumes of data at its disposal. The real problem, therefore, is not to find the next file, dataset or technical solution. It is about bringing existing information together so that it can play a role in the process.

The strongest counterargument to Sebastian’s approach would be to look for the bottleneck first in better stand-alone systems. His own diagnosis points the other way: Even powerful technology offers little major advantage if data between planning, financing, construction and operations remains isolated. The connection must therefore be established before, or at least in parallel with, the technical deployment.

Sebastian’s Perspective Explains His Approach

The academic

Sebastian holds a degree in business administration and teaches real estate finance at Irebs. He is also Director at the Center for Finance at the University of Regensburg and a Research Associate at the Centre for European Economic Research (ZEW). His research focuses include, among other things, indirect real estate investment, real estate finance, rent regulation and rental price indices. He is also a co-editor of the Journal of Real Estate Research and the German Journal of Property Research.

These areas of focus show that Sebastian does not view real estate solely as an individual construction or administrative property. His research connects capital markets, regulation and ongoing operations. This breadth is reflected precisely in his answer: The untapped potential lies at the transitions between the phases.

A second thought from the conversation

In another passage of the interview published in Handelsblatt, Sebastian names the introduction of negative interest rates as his biggest surprise. For decades, the zero lower bound formed a central reference point in theory and academia. His lesson from this is that economic modeling assumptions should not be treated like laws of nature.

At first glance, this may appear to have little to do with AI in the real estate industry. Both statements, however, share the same underlying attitude: Existing tools or assumptions do not automatically constitute the decisive limit. What matters is how they are connected to reality. In the real estate industry, Sebastian’s formula for this is: Connect data rather than merely collecting it.

What Sebastian’s Answer Establishes – and What Remains Open

The diagnosis is robust

Sebastian provides a clear assessment: The industry already generates enormous volumes of data. According to his observation, it generally uses these data in isolation. His proposed solution is better linkage between planning, financing, construction and operations – and the use of AI to unlock this untapped potential.

In this answer, he does not name any specific software products, individual AI models or efficiency gains that have already been measured. His statement should therefore be read as a decision on direction, not as evidence of a finished system. What the connection would technically need to achieve, how companies would organize it and what results it would produce remain open questions.

The question to ask about real estate projects

Sebastian’s assessment becomes practically useful only when you examine the data flow of a project more closely. The decisive point is not merely whether data is available at every stage. The more important question is whether information from planning, financing, construction and operations is actually available for the respective next decision.

This is where Sebastian’s strongest idea lies: The industry must link its existing collections of data within a shared workflow. Only then, in his assessment, can AI deliver the potential contained in the many isolated data sets. The issue is therefore not about selecting “the best AI.” It is about creating the connections on which meaningful AI can be built in the first place.