Artificial Intelligence: Boon or Bust for Property Managers?

This summer, Tesla CEO Elon Musk made headlines when he called artificial intelligence (AI) an existential threat to mankind, predicting that robots will one day do everything better than people, thus taking away their jobs. 

Building managers, be warned: you might be the first to go. 

A few weeks ago MetaProp NYC, a real-estate technology accelerator, announced an investment in Travtus, an AI platform for property management. Travtus is part of MetaProp’s incoming class, earning themselves a boost of resources and mentorship that Travtus founder Tripty Arya hopes will spell an Al launch in early 2018. 

Here’s how it works. Building managers and residents feed individual apartment work orders into Travtus’s AI software, which then interprets those orders in two ways: first, by comparing the work order to others that have been issued in the building, so as to alert the manager to issues that might extend across several units; and second, by comparing the work order in countless buildings citywide, helping the manager find solutions from beyond his purview. So if, for example, if several adjoining units suffer water damage, Travtus might inform the manager that rather than address each unit individually, it would be more cost-efficient to fix the facade. 

“Tripty has found a real niche in the market,” says Leila Collins, head of MetaProp NYC's accelerator and one of the figures involved in the decision to greenlight Travtus’s funding. 

Arya says she’s still working on pricing, but that when the program launches in early 2018, the cost will not be prohibitive. “I’ll put it this way,” she says, “pretty much every property manager will be able to afford it.” 

That’s in part because the AI model needs broad participation in order for the data library to become sufficiently powerful. After all, the library is “self-building” – the more work orders that go through Travtus, the richer the database it learns from. In some ways then, Travtus is similar to social media platforms like Facebook: offering a service to its users while mining value from increased user activity. 

But for co-op and condo boards who might blanch at the thought of feeding a quasi-public database the nitty gritty on every insect fumigation in the building, Arya says those skeletons will be kept safely in storage. “There’s a lot of data security built into the system,” she says. “None of the data has any indicators on it,” meaning the data is encrypted, and the building, tenant, and apartment are not logged alongside work orders in Travtus’s database. 

Arya still has work to do to get the system ready for launch. One issue will be coding the system so that it recognizes the difference between causation and correlation. By way of example: a lot of poorly maintained apartment buildings might have problems with both mice infestation and broken stairwell lights, whereas well maintained apartments might have neither. But that doesn’t mean mice complaints should be addressed with light installations. 

“It’s a work in progress,” Arya says. 

If this AI becomes an everyday reality of property management, it will join a growing list of tech innovations that are catching on with co-op and condo boards. This one might be a money-saver. With property managers doing more cost-effective work at shorter hours, fees might plummet. 

“It’s not going to replace the property manager,” Arya says. “It’s just going to optimize their role. Instead of focusing on 40 units, they’ll be able to look at 400.”