Proptech at Scale: How Real Estate Teams Are Vetting AI Without Getting Overwhelmed

The Short Version

• Real estate owners and operators are testing a growing number of AI and proptech tools as they work to determine which technologies deliver measurable value across their portfolios.
• Technology is creating efficiencies in areas like leasing, legal review, facilities management and data analysis, while generative AI is becoming more embedded in daily workflows.
• Companies are developing formal processes — including training programs, AI councils and internal guidelines — to evaluate new tools, manage risk and scale technology adoption more effectively.

The Industry Is Sorting Through AI Hype To Find Real Value

The real estate industry is still figuring out what AI works, what doesn’t and how best to implement and scale technology across portfolios. And the options are overwhelming; vendors pitch property owners with new tech on an almost daily basis. “Owners and operators are really starting to treat technology as a strategic lever, whether it’s building internally or delivering through existing partners,” said Tanger senior vice president of technology Talia Fine. The most significant challenge is navigating the gap between hype and reality, she said.

Landlords and retailers, from small to large, are wrestling with the question of what AI means for their business, added Jim Dillavou, principal and co-founder of Paragon, a grocery-anchored shopping center developer and owner with properties on the West Coast. “There are a million different beta tests occurring in the market right now as people figure out how to optimize for this,” he said.

For example, some companies are experimenting with AI models that allow them to create mirror images of cities with the same consumer demographics and spending patterns in order to predict sales volumes. What happens when a retailer puts a store in that city? How does that performance change when they move that store to different possible locations within the city? “That shows the extent to which these tools can be used in the future,” said Dillavou.

MORE FROM C+CT: Digital Twins Bring Real-Time Intelligence to Retail and Real Estate

Early AI Use Cases Are Delivering Operational Efficiencies

From a practical perspective, AI is creating efficiencies from facilities management to leasing, accounting and marketing. Paragon recently tested the use of legal AI bots, including Harvey and Paxton, in the recent purchase of a shopping center in Southern California. The bots did much of the legal work, including purchase and sale review, lease review, survey and title review and title endorsement recommendations.

“AI is all about prompting, so it is critical to be prompting it properly and lawyers are still needed in transactions,” Dillavou said, noting that AI offers efficiency by instead helping with more task-oriented work.

AI Tools Move From Experiment to Daily Workflow

Early adoption of AI and large language models focused on restructuring workflows and improving communication with vendors. The industry continues to innovate.

Tanger, for example, embedded an AI assistant into its ecosystem, and the team uses the tool for tasks like extracting and summarizing documents, generating charts with citations and conducting general research. Tanger also is leaning into predictive analytics. Initially, the firm used machine learning to forecast sales and traffic, and now the tech incorporates generative AI. The company recently launched a new deal history report that combines lease‑level financial data with detailed clause analysis. The platform produces AI‑generated summaries of contractual language and provides a confidence rating to help users assess summary reliability. “We think that’s going to be a really powerful tool for our leasing, legal and other departments, as well,” said Fine.

A new version of Placer.ai, set to launch in the second quarter, will allow users to ask Placer.ai specific questions. For example, a shopping center owner looking to lease a 3,000-square-foot vacancy can ask for recommendations for retailers that fit that square footage and can drive a certain amount or range of sales. After Placer.ai provides suggestions, the user could prompt it to create a proposal for a specific tenant that highlights the attributes of the shopping center and local market.

The new Placer.ai will bring together public and private data, identify patterns and present information in a way that is easily understandable, said NewMark Merrill Cos. president and CEO Sandy Sigal, who also was an early investor in Placer. That’s “hugely powerful” for an industry driven by data, he said.

Operations Technology Is Improving Oversight and Reducing Risk

Real estate companies also leverage tech to add value and manage risk. For example, JLL has maintenance management system that uses a lease abstract tool to determine what work is a tenant’s or the property manager’s responsibility and that tracks what work is being done. Vendors are required to check in and out of the system when working at a property. “That’s a game changer because you know who’s in your building, you know what they’re doing and you know that your tenants are doing what they’re supposed to in maintaining their space,” said JLL retail property management director of operations and senior vice president Barry Wood.

Companies Are Building Formal Processes To Evaluate AI Tools

A company needs to make sure a technology can deliver on the organization’s objective for that particular tool, and so a challenge arises in evaluating and testing before implementing and scaling.

Without a formal structure to evaluate new technologies, a company could chase them down different rabbit holes, noted Sigal. Not only is that expensive, but resources spent on tech that doesn’t work can erase the productivity gains it was supposed to deliver in the first place, he said.

NewMark Merrill’s strategy is threefold. It hosted an AI workshop for all employees so all could build on the same base skill set and understanding. The company’s C-suite and department heads also populate an internal AI council, which this year aims to implement two innovations for each of its departments. Third, the company developed internal AI and technology guidelines that outlined “dos and don’ts,” such as testing agentic AI tools in a way that ensures data privacy.

“The best thing that companies and individuals can do is focus on building that AI muscle,” said Fine. That includes developing technical capabilities, establishing governance frameworks, understanding the risks and upskilling employees, she said. Tanger also has a formal AI council that includes individuals from every business line. It comprises three subcommittees: Innovation & Development, Education & Impact and Risk & Compliance.

Technology seems to move so fast that by the time teams can evaluate, test and implement, a new iteration or generation has become available. “AI is moving at warp speed,” said Wood, “and the challenge that we have is we want to make sure what we’re implementing is right.”