The real estate industry, which has historically been reticent to adopt new technologies, is embracing artificial intelligence in several forms, ranging from enhancing visuals and engaging clients to summarizing strata documentation and improving realtor workflow.
The current applications of AI in real estate include finessing property renderings, communicating with potential buyers and sellers, streamlining transactions and regulatory compliance, analyzing complex documents, boosting administrative productivity and predicting future listings.
Key goals are to save time, cut costs, eliminate human error, identify new business opportunities and drive sales. But AI also brings challenges of its own around human decision-making, raising ethical questions about accuracy and the need for proper checks and balances.
Experts say it’s only the beginning, and that AI will soon be able to achieve more grandiose aims, such as optimizing housing policy and urban planning to make the right decisions faster for positive impacts on communities. AI could even play a major role in addressing the shortage of affordable housing.
“If we are actually able to use this technology to solve some of the smaller problems that we have in the industry, what is then the role of it to be able to solve some of these bigger challenges when we think about development and the infrastructure that goes around it?” said James Innis, president and COO of Sutton Group.
“Those are the meaty problems that are really roadblocks from a real-estate standpoint, particularly here in Vancouver.”
Early applications include virtual staging, chatbots, data analysis
AI can generate or enhance the visual media that accompanies pre-sale marketing.
“On the very basic level, AI has drastically improved the realism and quality of renderings when they are taken to final high-res,” said Dan Hashemi, CEO of LNG Studios.
“AI just adds that final little bit of touch to the plants, the ground, the materials on the building, or even converting 3D people into 2D people and having that hyper-realism.”
He also said the technology allows his team to explore different angles, times of day and “concepts around how we want to tell the story of the image, before our artists even get started.”
Another early application is the use of chatbots to interact with the public. These bots can book appointments, schedule showings and follow up with past clients.
“It doesn’t really feel like a robotic sort of an experience because of the advancements of AI,” said Hashemi. “It still feels very much like you’re speaking to someone who understands your needs and is able to really guide you.”
Other local companies focus on data analysis. For example, The Real View uses technology to quickly analyze strata document packages for prospective buyers of condos and townhomes. These documents can include council meeting minutes, financial and insurance documents, bylaws and structural engineer reports.
“Especially in Vancouver, it’s quite a hot market for real estate from time to time,” said principal and co-founder Andrew Armstrong. “If you’re looking at a place on Sunday and you need to have your offer in by Tuesday, this document package in some instances is 1,800 pages long. That doesn’t really give you the time to look through it all yourself and give yourself a good understanding.”
The Real View, which has about 500 realtor-clients as well as a national partnership with Sutton Group, uses RAG—or retrieval-augmented generation—to analyze strata document packages. It takes about 15 minutes to produce a comprehensive report on key aspects of the building, ranging from its financial health to its pet policies.
Through optical AI, the company can analyze these documents, even if the PDF quality is poor. Because there is no standard format for strata document packages, and because The Real View’s technology was trained for 18 months on many thousands of iterations, it can reduce human error.
“The human error aspect is that most people are looking in specific places for information,” Armstrong said. “If they don’t find it there, they assume it doesn’t exist, whereas we look at every page of every document for every answer.”
Another local startup is Saige, which aims to increase a realtor’s listings through AI. The company has a proprietary algorithm that identifies homes that are likely to be listed, whose owners can then be targeted through direct mail and other outreach.
“Out of a neighbourhood, what are the homes that are likely to list in, let’s say, the next six to nine months?” said founder Mehdi Aghaei. “We narrow down [a realtor’s] searches to those homes that are likely to list, rather than every home that is out there.”
Saige looks at about 200 different features of homes for sale and feeds this data into its model to predict similar homes that may get listed in the near future. The company launched in 2023 and has about 60 teams of realtors using its platform.
AI can help with AML compliance and closings
Sutton Group’s Innis said his brokerage will soon roll out AI tools around regulatory compliance and deal closings.
In compliance, AI can be used to assist with anti-money-laundering (AML) disclosure mandated by Fintrac, the federal government’s financial intelligence unit. Fintrac requires realtors to complete and submit a set of forms to their brokerage related to KYC, or know your client, which can then be audited by the agency. Currently, there is a limited set of tools, but AI will soon provide a collection mechanism that simplifies the process.
“Everyone wants to be compliant, so how do we make that easier?” said Innis. “You’ve created a data set of individuals and transactions. You can then start to use AI models directly on top, scanning for particular transactions, asking questions, analyzing things from a managing-broker standpoint, all with a view to mitigate your risk.”
Real estate closings can also benefit from AI, he said.
“It can go through paperwork and say, ‘This hasn’t been done correctly, here’s what you need to do.’ You can ask it questions, such as, ‘How do I manage this, how do I fill that out, what am I missing?’ It provides that back-and-forth which is a role that is typically filled by the managing broker. These are busy folks.”
AI has limitations, brings ethical implications
Experts say AI has constraints like any other nascent technology. For example, members of the public can potentially be misled by artificial images of real estate, a problem that is being addressed by some local real estate boards for consumer protection.
There are also risks from drawing normative suggestions or conclusions from AI. One should be wary of using AI to make value judgments, said Armstrong of The Real View. For example, AI may not be able to draw proper inferences regarding a strata’s financial health.
“If that happens to be a mega-tower out in Brentwood, for example, $400,000 is really not enough money to have in a contingency reserve fund, likely,” he said. “By putting your stamp of, ‘Yeah, that’s a great amount’ because you’re just basing it off general information, that’s where you can get yourself into trouble.”
AI also needs human monitoring and spot checks to avoid blind reliance on imperfect technologies. This is especially true where an AI model is fed incomplete or outdated inputs.
For example, AI can be used to accelerate municipal permitting. But if the city lacks an updated official community plan or zoning bylaw, or lacks updated infrastructure plans, these deficits will limit a city’s ability to leverage AI and chatbots to provide specific feedback to applicants at the site level.
AI “gets me a lot of the way there because I’m able to have a certain probability of success,” said Tegan Smith, CEO of Channel Consulting. “I’m able to say it’s probably correct, and then I can go and verify. But you have to actually have the expertise to know what to ask and what to verify. I think the critical thinking and human involvement is going to continue to be important.”
Eventually, AI could be used on a macro scale, said Sutton Group’s Innis.
“What if you started to ask an AI model the question, ‘What do you think the housing target should be for Vancouver?’ Or, ‘Should we do the development?’ Or, ‘What should the limit be on mass timber?’
“That’s my big aspiration for AI,” he said, “but we’ve got to do the hard miles now with some of these smaller pieces.”
