Made to measure: why real estate businesses are using AI to build their own tools

Key points

• Real estate investors and agencies are using AI to build their own tools rather than buy them.
• Firms typically build alongside their core systems, not instead of them: investor reporting, lease management and deal screening are common starting points.
• The first step is mapping existing workflows to see where AI makes the biggest difference.
• Building with AI still needs proper design, testing and governance.

Green Street News looked at how real estate businesses are moving from buying software to building their own tools with AI, from data platforms to virtual investment committee members. The feature draws on Segro, Landsec, Instruction, DTRE and Tidebridge.

It starts with Segro's "taxi pack", which lets asset managers pull up and query occupier data on the way to a meeting, and the distinction Segro's chief information officer Rich Corbridge draws between "small AI" for everyday tasks and "big AI" that transforms workflows across a business. The first step, the article says, is mapping existing workflows to see where AI makes the biggest difference. Investor reporting, lease management and deal screening are given as areas where firms are building their own tools, usually alongside existing software rather than replacing it.

The article cites DTRE's new DXTRE data platform as an example from agency, and describes Tidebridge as having partnered with DTRE on its development and as advising other real estate businesses on developing custom systems with AI. Oli Farago argues that it makes little sense to replace core specialist systems that would be complex and time consuming to recreate, but that the way companies interact with those systems is changing. The piece closes with a note of caution: the RICS professional standard on the use of AI, and the risk that building with AI without proper design and testing produces poor code and security risks.

Our view. We agree with the caution as much as the opportunity. The right answer is rarely to replace a core system such as a property management or valuation package. It is to build the layer your teams actually work in, connected to those systems, designed and tested like any other production software, and owned by you.