How One Brokerage Uses AI to Find Deals Faster
Diamond Real Estate Group isn't using artificial intelligence to churn out listing descriptions or automate inboxes. Instead, the Northern California brokerage is quietly embedding AI into the mechanics of how deals actually get done.
For CEO Matthew Martinez, the real value of AI isn't in flashy outputs—it's in building a system that can surface the right opportunities faster than any human could on their own. At a firm of 10 working across commercial, luxury residential and vineyard properties in Sonoma, Napa, Marin, Solano and San Francisco counties, that edge can make a measurable difference.
Artificial intelligence has become a core part of how the firm searches for and analyzes properties on behalf of buyers and sellers. Martinez has built custom workflows using Anthropic's Claude large language model alongside automation tools from German software company n8n, creating a system tailored to the firm's internal data and processes.
"The beauty of AI is I can upload all sorts of market data, tax records, [and] do the market analysis that might suit our clients' needs that might not be on the market," he tells GlobeSt.com.
"It is the combination of AI with actual workflow systems, internal data, and repeatable operating processes, so it affects how deals get worked in the real world."
Rather than relying on AI as a standalone tool, Martinez is focused on integrating it into repeatable business functions. The result is a set of AI agents designed to streamline property discovery by rapidly analyzing large datasets against specific client requirements.
He describes these agents as tools that "understand the business" and which "can parse through data a hell of a lot faster than I can." But he is clear-eyed about their limitations.
"I take the outputs and use my own brain to do what I will with it," Martinez says. "I take the outputs and use my own brain to do what I will with it. I put in parameters, obviously. It can't think on its own. I know exactly what I'm looking for. Then it uses all the data I've uploaded to give me its best answer."
In practice, Martinez sees the technology less as AI in the popular sense and more as a highly advanced filtering system.
His best description is "an extremely efficient search engine," one that matches client-defined parameters against a large pool of uploaded data. Those parameters—ranging from location and asset type to financial and tax considerations—allow the system to quickly surface potential opportunities. From there, human brokers refine and evaluate the results.
The setup is relatively affordable, costing between $100 and $120 per month for the firm's current usage. The bigger hurdle isn't cost but adoption.
Getting agents comfortable with the system has proven uneven. For some, it's "a little bit of a learning curve." But "it's not for everyone. For some people, it's way too foreign," according to Martinez.
Even for Martinez, refining how to interact with AI tools can require extra steps. At times, he turns to OpenAI's ChatGPT to help structure prompts more effectively for Claude, underscoring the practical complexity of working across multiple AI platforms.
"Getting it set up initially is a big task," says Martinez. "Tailoring it to your business is probably the biggest initial challenge. "Tailoring it to your business is probably the biggest initial challenge."
That upfront effort, however, is what ultimately determines whether AI becomes a novelty or a meaningful operational advantage.
Source: GlobeSt/ALM