AI in multifamily maintenance is still early innings—mostly reactive, not yet predictive. GlobeSt. explores how HappyCo customers use smarter documentation and visibility to catch problems before costs compound.
By Richard Berger
Many apartment industry operators seem to be rushing toward artificial intelligence for just about everything, from seeking solutions, greater efficiency, to potential cost savings.
But in this technology’s “early innings,” some are realizing that AI is still mostly reactive, responding to problems faster than preventing them.
Jeff Derricott, director of operations – IT & Process Strategy at Thomas Safran & Associates (TSA), based in Los Angeles, has shifted his operational software to deliver results.
“Previously, it was time-consuming and difficult to gain visibility into performance metrics surrounding preventative and active maintenance,” he told GlobeSt.com.
“Benchmarking property performance, identifying service gaps, tracking asset degradation — these capabilities don’t just inform strategy, they help us intervene before inefficiencies compound into costs. When you're managing large affordable housing operations, visibility into maintenance operations is crucial.”
TSA manages about 70 properties in Southern California, totaling nearly 7,000 units, curating homes for low-income families and individuals.
“Having in-depth portfolio-wide visibility allows us to fulfil commitments to our residents with greater ease and more focus.”
With operating expenses putting pressure on multifamily performance, operators such as TSA are taking a closer look at controlling maintenance costs, finding that it is one line item they can manage.
Maintenance can be hard to benchmark across properties, according to Ben Nowacky, president of software provider HappyCo, who spoke at the recent NMHC's OPTECH Conference.
“Service gaps aren’t always visible until they become costly. Equipment degradation doesn't announce itself,” he told GlobeSt.com. “When documentation is inconsistent or incomplete, operators may end up dispatching techs for repeat visits or missing early warning signs—inefficiencies that add up over time.
"Much of what technicians know never makes it into a system—it leaves with them at the end of a shift or when they turn over. The goal is to turn everyday repairs into data that informs bigger decisions.”
The future of maintenance is moving toward predictive and proactive maintenance, according to Nowacky.
“You want to identify the issue and have signals about a problem before it ever happens,” he said.
In fact, Jordan Arend, system strategy lead for Midwest apartment operator Price Brothers Management, told GlobeSt.com that detailed documentation solves issues the first time.
“When techs capture what failed and how they fixed it – photos, voice notes, everything – we're not dispatching them back for the same issue," he explained.
"Better yet, we now know which technicians to send based on skill set, proximity, work order complexity, and cost efficiency. All that intelligence is accessible in real time.”
Nowacky points to real cases where operators missed signals buried in their own data. In one example, a roof leak led to mold remediation—not because the damage wasn't visible, but because work orders came in about “bad smell” and “organic matter” rather than water.
“The dots were there; no one connected them,” he said.
HappyCo analyzed data across its 5.5 million units and found little correlation between resolution time and resident satisfaction. What mattered more—communication and proactive updates.
"A lot of teams are focusing on the wrong numbers, handing out bonuses to techs based on speed,” Nowacky said.
“They're incentivizing techs to solve work orders quickly—but that doesn’t lead to resident satisfaction.
"Having properly stored documentation can be used to troubleshoot issues or even prevent them from occurring, rather than simply rushing through the work order.”
In a recent technological step forward, HappyCo has introduced voice-assist capabilities, further improving efficiency in the daily work-order process.
AI-powered voice captures and transforms field observations into automated workflows and predictive intelligence, based on more than a decade of service records across 5.5 million units, accumulating more than a billion maintenance interactions.
Technicians dictate completion notes naturally in the mobile app. The system extracts what was fixed, how long it took and which materials were used – creating structured, searchable records in real time.
Lindsey Henn, HappyCo PR
E: lindsey.henn@happy.co
M: +1 (626) 893-42298
