Stop Buying Chatbots: The Case for Large Action Models in Real Estate

The real estate industry is currently in the middle of an “AI-washing” epidemic. Everywhere you look, legacy software providers are slapping a generic ChatGPT-style chatbot into their sidebars and calling it a revolution.

But for the modern brokerage, these passive chatbots are often more of a distraction than a tool. According to research from McKinsey, over 70% of independent actions actually fail to reach the intended goal.

In the fourth episode of our SERHANT. series, Chief Experience Officer Ryan Coyne explains why the world’s most innovative brokerage isn’t interested in chatbots. Instead, they are building their future on Large Action Models in real estate.

The Problem with Passive AI

The fundamental flaw in most real estate AI today is that it is passive. It waits for the agent to ask a question, and then it “spits out” a wall of text that the agent then has to copy, paste, format, and send themselves.

“Passive AI that’s just waiting for you, and only spits something back out that you then have to go produce something on your own, isn’t enough,” Coyne explains. “Real estate doesn’t need another chatbot. It needs a system that actually goes and does things and achieves tasks for you.”

If the AI doesn’t actually remove a step from the agent’s workflow, it hasn’t truly solved the problem. It has just moved the “busywork” from one window to another.

The “Sous-Chef” Philosophy: Automate the Repeatable

At SERHANT., the goal of AI isn’t to replace the agent, but to amplify their humanity. This requires a philosophy Coyne calls “Humanity Amplification.”

He uses a culinary metaphor to describe how SERHANT. utilizes Rechat’s AI: “We decided very early on that it was the best thing to do to have almost like sous-chefs in a restaurant kitchen that are inspecting plates before things leave the kitchen, and just arranging or tweaking it just so, so it’s perfect when it finally goes out to the table.”

By automating the repeatable, administrative “prep work,” the agent (the Head Chef) can focus entirely on the exceptional: the client relationship, the negotiation, and the emotional intelligence required to close a $50 million deal.

“If we keep humans in it, you end up keeping things more human, you end up keeping people active and engaged and the quality very high,” says Coyne.

Lucy: Bringing Large Action Models in Real Estate to Life

The real estate industry is currently in the middle of an “AI-washing” epidemic. Everywhere you look, legacy software providers are slapping a generic ChatGPT-style chatbot into their sidebars and calling it a revolution. While many focus on basic chatbots, the true future belongs to Large Action Models in Real Estate.

This is where Lucy, Rechat’s proactive AI assistant, separates itself from the “chatbot” crowd. Lucy doesn’t just talk; she executes as a true Large Action Model in real estate.

When a listing status changes in the MLS, Lucy doesn’t wait for a prompt. She proactively generates the “deliverables”—the newsletters, the social media posts, the property presentations, and the email campaigns. She prepares the “plate” for the agent to review, approve, and send.

Because Lucy is integrated directly into the Rechat Operating System, she has access to the agent’s CRM, their marketing suite, and their transaction data. This allowed Rechat users to generate significantly higher revenue in 2025 by focusing on high-value activities while the Action Model handled the noise.

Beyond the Hype: Building for 2030

As we move toward a more tech-driven market, the gap between brokerages using basic “chatbots” and those utilizing true action models will only widen. Meaningful AI in real estate should be measured by the deliverables it produces, not the words it writes.

“I like that Lucy is something that’s actually able to produce deliverables and take actions on your behalf,” Coyne concludes. “It minimizes the amount of effort that you have to put in and gets you back to your clients and back to your life.”

Watch Episode 4 of our SERHANT. series to hear Ryan Coyne’s full breakdown of the Large Action Model framework and why SERHANT. is betting on “Psychology-First” AI (or start from the beginning of the series: Ep1: How Ryan Serhant Uses Targeted AI Marketing to Drive 144% GCI Growth).

Download our comprehensive 2026 State of AI & Real Estate Marketing Report to discover the power of Agentic AI and Large Action Models.