Walker Webcast: McKinsey’s Aditya Sanghvi Discusses Why and How Agentic AI Can Be Tailor-Made for Rental Housing

Artificial intelligence, particularly the agentic kind, could help rental housing owners and operators automate tasks, coordinate maintenance and improve resident retention. But getting there involves more than installing a new technology.

“I would put out there that rental housing is perfect for AI,” said McKinsey & Co. senior partner Aditya Sanghvi, a leader of the firm’s global real estate practice. Sanghvi spoke at Walker & Dunlop’s recent annual convention in Sun Valley, Idaho, in a session broadcast Aug. 12 as a Walker Webcast.

During the session, Sanghvi discussed why rental housing is well-suited to agentic AI, why adoption remains slow and what CEOs can do to get more value from the technology.

Rental Housing Is Built for Agentic AI

Unlike generative AI, which responds to human prompts, agentic AI is given a goal. The agent then plans and executes the tasks needed to achieve it. Multiple agents can also coordinate specialized processes.

Sanghvi said that the technology perfectly matches property management, as the industry relies on numerous relatively simple tasks that become complicated when residents, property managers and vendors intersect. The result can be “dead zones,” the unfulfilled maintenance tickets, unanswered calls and missed follow-ups that can erode the resident experience.

Agentic AI can help fill those gaps. In a maintenance situation, for example, one agent could triage requests, another could coordinate vendors, and another could handle scheduling. A separate agent could identify recurring problems and improve future decisions.

Agentic AI could also improve lease conversions, renewals, collections and other repetitive processes. Over time, agents can learn from repeated decisions and become more effective.

This doesn’t mean that humans are eliminated.

“Humans are essential for the loop of every agent AI process,” Sanghvi said. Humans still outperform agents on many tasks. Also, agents can’t be held accountable.”

Adoption Remains Slow

Despite the potential, Sanghvi said only 6% of companies report a positive financial impact from AI.

One obstacle is deployment. Generative AI can be installed for specific tasks, but agentic AI requires companies to rethink how work gets done.

Data quality, or lack thereof, is another major issue. A Gartner report found that 60% of companies have halted AI efforts because of poor data quality.

“The data in real estate is terrible,” Sanghvi said, noting that property management systems, spreadsheets and CRM databases often don’t connect.

Organizational issues add another problem. Leaders frequently delegate AI to IT rather than taking ownership themselves. Also, real estate companies may lack the expertise needed to implement and maintain the technology.

Change management is equally important.

“Unless people know how to use the tools, unless they see other people using them, unless their rewards and consequences are set up against it, they won’t use it,” Sanghvi said.

What CEOs Need to Do Now

Sanghvi recommended what McKinsey calls a “rewired” approach: Tie AI efforts to measurable value, reimagine the work and build change management into the process from the start.

He suggested focusing on specific “domains” that offer clear goals and key performance indicators. For instance, focus on a domain covering leasing renewals, which directs companies to clean and organize tenant data while also testing an AI-driven process. The result is a feedback loop that gets the job done and helps validate and improve results.

Data governance and guardrails are also critical. Agents should operate within defined boundaries, with human oversight for important decisions.

Sanghvi said that rental housing’s repetitive, interconnected workflows make it particularly well-suited to agentic AI. Success depends on redesigning work, improving data and creating continuous learning loops.

“Rental housing is built for this,” he said. “Those who win will be the ones who redesign work the most, and who have the data, and then start these learning loops that make every decision better.”

On-demand replays of the August 12 Walker Webcast are available through the Walker Webcast channels on YouTube, Spotify and Apple. Subscribe to get invites, replays and articles for new Walker Webcast episodes every week.