Apartment Industry Incorporates AI Into Operations

Apartment management has come far enough so that employees shouldn’t have to organize their email inboxes first thing every morning.

So says Stacy Holden, AppFolio, speaking on an industry panel at the National Multifamily Housing Council’s OpTech Conference this week in Las Vegas.

Artificial intelligence is here “to do that for you, arranging emails in a way that the employee knows what’s ahead for them and allow you to work on what you specialize in,” she said.

AI 101 was the topic of the panel moderated by Sharon Wilson Géno, NMHC, with panelists James Scott, Massachusetts Institute of Technology (MIT) Real Estate Innovation Lab; and Amy Glassman, Ballard Spahr LLP.

Scott raised the issue in many attendees’ minds about “when AI was going to take over” everything.

He said it could happen in 2 to 200 years, “but we don’t know,” leaning toward it likely won’t be in two years and that today, business is using it in a reactive [not proactive] way.

For now, AI is something industry supplier partners and owners and managers are working together on to “meet their customers where they are,” Holden said.

She said AppFolio has working on AI for five years and currently 85 percent of AppFolio’s customer base is using it in some form.

“The ways it can be used is increasing incrementally faster and faster,” Holden said.

“You might not know, but your employees are using it. You need to know how they are using it.”

Scott mentioned a student he had who had to summarize a 300-page real estate legal document – a project that typically took about 15 hours.

“Using ChatGPT, he was able to get through the assignment in about 20 minutes.”

Holden said it’s crucial that AI users know where their data is coming from as AI’s general knowledge continues to grow and improve.

Conference participant Itamar Roth, CEO of Colleen AI, said the reliability and accuracy of AI is only as strong as the magnitude of the data set or scale of units it has trained with.

“Therefore, it’s essential to select a solution that has a proven track record and served a large number of residents over time. Applying this to the multifamily sector equates to hundreds of thousands of units over at least one year.

Additionally, he said there is also a challenge for AI to operate in an environment with a lot of human interactions.

Therefore, industry solutions that come with a promise of solving all property management challenges are impractical.

Some new companies might take shortcuts by relying solely on generative AI, which is trained on unvetted Internet data.

“This can lead to errors, and for critical services like rent collection, there should be zero tolerance for mistakes.”