Artificial intelligence could generate returns of 20–30% when properly integrated into property operations. However, fragmented systems, inconsistent data and organisational barriers are limiting its potential.
According to Economx, Hungary remains behind countries such as the Czech Republic and Poland in property-sector digitalisation. The challenge is less about accessing AI technology and more about connecting existing systems and organising data effectively.
From Construction to Operations
At the AI Summit, Zoltán Kalmár, founder of Proptech Hungary, highlighted the property industry’s continued reliance on Excel, where the same information can appear in different files and formats. Fragmented data makes it difficult for AI to turn information into useful business knowledge.
AI can support the entire property lifecycle, from design and construction to operations. In construction, this could be particularly valuable as around 70% of projects experience some degree of delay and nearly 20% exceed their budgets. AI can help predict risks and subcontractor problems and support faster cost checks.
The Biggest Opportunity Is in Operations
Property operations may offer the greatest potential because this stage lasts much longer than design or construction and generates substantial volumes of data. Combining AI with IoT and BIM can help optimise energy consumption, predict maintenance needs and improve building operations.
This long operational lifecycle also explains why a significant proportion of proptech businesses are targeting this part of the property market.
However, technology alone will not deliver these benefits. A major barrier is integrating AI into company structures and creating a culture where employees can experiment, make mistakes and learn, Kalmár stressed.
Many property companies already use project management, ERP, CRM and other software, but these systems often operate separately. AI cannot automatically create a unified decision-support system when the underlying data remains fragmented across disconnected platforms.
The priority is therefore to connect existing systems and embed AI into established workflows rather than simply creating another standalone AI function. AI tools are already moving beyond question-and-answer tasks.
Multimodal AI can also work with text, plans, photographs and spreadsheets simultaneously. The next development could be AI agents capable of carrying out several stages of an entire workflow.
Start Small, Then Measure
Companies can begin with smaller processes that can be redesigned within days or weeks. Results can then be measured through time saved, working hours released, fewer errors and direct financial savings.
The technology is already available. The key challenge is turning AI from an individual experiment into an integrated part of everyday property operations.