AI Agents Finally Automate Commercial Real Estate Ownership Research

Commercial real estate ownership research has long been a manual bottleneck in an otherwise digitized brokerage workflow. The reason, according to Dan Mosher, CEO and Co-Founder of DealGround, is structural: property records are scattered across 50 states and thousands of counties, each with its own update schedules, formats, and accessibility. Until recently, no technology could efficiently connect the multiple steps required to trace an owner from an LLC filing to a phone number and email address.

County records update on different timelines—some within hours of a transaction, others take weeks. State Secretary of State filings vary in format, accessibility, and completeness. Because most commercial properties are held in LLCs or trusts rather than individual names, finding the actual person behind a property requires navigating multiple distinct systems in sequence. “Every state is different. Every county is different,” Mosher says. “There is a fragmentation of the properties because they’re all managed locally.”

The process involves at least three separate functions: identifying the LLC or trust that holds the property, piercing that entity to find the individual behind it, and locating current contact information for that person. Each step draws on different data sources with their own structure, access rules, and update cadence. The accuracy of each is informed by the region.

According to Mosher, the core problem was never that technology could not handle any single step; it was that no technology could handle all steps in sequence. Filtering down to the right set of properties for a particular client was possible. Building a tool to search Secretary of State filings was possible. Building a tool to look up contact information was possible. Connecting those three functions into a single automated workflow was not. “You could probably build technology in each of the three steps, but you could never chain all the steps together previously,” Mosher says. “Now you can chain them all together, and that’s what we offer, which has never been done before.”

Mosher attributes the change to AI-driven agentic processes, systems that execute multi-step workflows autonomously and move from one function to the next without human intervention at each stage. He describes this capability as viable for roughly the past year, which helps explain why ownership research remained manual even as other parts of the brokerage workflow were digitized.

Mosher says brokers doing active prospecting can spend 10 to 20 hours per week on ownership research alone. He describes speaking with customers who were logging 15 hours a week on this work before adopting an automated approach, and who now accomplish the same output in 15 to 30 minutes.

The time cost of manual ownership research represents a structural constraint on how many productive conversations a broker can have. If a broker spends 15 hours per week identifying owners and tracking down contact information, those are hours not spent on searching and filtering for properties, calls, pitches, or deal development. Mosher frames this as a reallocation of productive capacity, not simply a convenience upgrade.

Accuracy compounds the problem. Property owners managing multiple assets through separate LLCs tend to change phone numbers frequently and maintain multiple email addresses. A manual research process that takes days or weeks to complete may produce contact information that is already outdated by the time a broker attempts to use it. For brokers whose income depends on reaching owners before competitors do, stale contact data does not just waste time—it costs deals.

DealGround’s platform replicates the manual research process but executes it through AI-driven agents that can run multiple ownership lookups simultaneously. A broker can submit 100 LLCs at once, and the system works through Secretary of State filings, identifies the associated individuals, and retrieves current contact information without the broker managing each step. “We replicate the manual process today, so it’s not so much different, but we do it much faster because it’s all AI agentic initiated and executed,” Mosher says.

Mosher describes a case in which a broker searching for land parcels in Texas, where Secretary of State filings were incomplete, used DealGround to surface an owner name, email address, and phone number that manual research had failed to produce. “The fact that you’re able to discover this, this could be the difference between no deal and a deal,” the broker told Mosher, because the parcel had been unreachable through conventional methods.

Mosher notes that DealGround is not the only platform addressing this problem. He positions the company’s advantage around accuracy and freshness of data, running ownership lookups on demand so results reflect current information rather than a static snapshot. Mosher says DealGround is currently running at about 95% accuracy in the data it extracts from documents.

The barrier to automating ownership research was always the fragmentation of steps across disconnected systems. Now that AI agents can chain those steps into a single workflow, the bottleneck that constrained commercial real estate prospecting for decades is addressable in a way it was not before.