How This Investor Used AI to Bypass Brokers to Buy an Office in Los Angeles
A South Korean family office, managing investments for a semiconductor manufacturer, recently zeroed in on an office building in Los Angeles using an AI tool that let it sidestep real estate brokers entirely. The approach cut through weeks of back-and-forth negotiations and agent recommendations, delivering objective analysis on dozens of properties in days. This deal marks one of the early signs of how wealthy families are quietly reshaping their commercial real estate strategies with technology.
Sang-Hoon Kwon, CEO of Neotis—a South Korea-based producer of micro-bits for printed circuit boards—leads the family office behind the purchase. With business expanding in the U.S., Kwon sought a foothold amid a downturn in the office and commercial property sector. Lacking local market knowledge, he first turned to Korean agents, but their conflicting advice and perceived biases—promoting properties tied to their existing clients—pushed him toward AI.
He turned to Diald AI to evaluate options in the Los Angeles area. The tool allowed him to run analyses on 40 to 50 properties in about a week, generating clear benchmarks on pricing and value without human sway.
"The AI didn't have the bias option, and it was very quick," Kwon told GlobeSt.com, noting how it freed him from obligations to meet agents or chase their preferred listings.
From there, Kwon narrowed the field to three properties in the area, using the AI's data to set offer prices. Negotiations with sellers lasted two weeks, landing a deal close to his top choice at a benchmark-adjusted price. The entire process—from screening to near-close—took roughly one month, a fraction of the time spent on pre-COVID efforts with brokers that fell apart when the pandemic hit.
Steven Song, founder and CEO of Diald AI, sees this pattern repeating among family offices. These investors often run lean teams but chase complex deals against bigger players, so AI expands their analytical reach without adding staff.
"Family offices look at us as a no-brainer way to increase the ability to analyze more opportunities without increasing head count," Song explained.
Song highlighted Kwon's case as typical: without U.S.-based analysts, the family office screened 40 LA properties in a day, produced investment memos and flew in for a 24-hour site visit before making an accepted offer. About 20 to 25 of Diald's 84 platform users are family offices, drawn by the tool's knack for turning public data, local signals and policy context into digestible reports. One report flagged a property as 27% overpriced, guiding smarter bids.
Other family offices have used the platform for scenario planning, like testing office, hotel or residential conversions on a municipal sale.
AI adoption in commercial real estate remains in embryonic stages, though emerging cases like this point to future upheavals in how deals get done. Family offices value the broker-bypass not just for neutrality but for slashing time spent on repetitive conversations.
As Song put it, serious users have moved from piloting tools to workflow staples, with his platform's client base surging eightfold in two months. For investors like Kwon, who is eyeing more U.S. buys totaling $10 million initially, the edge is clear: faster conviction in a crowded field.
Source: GlobeSt/ALM