How a Real Estate Broker Built His Own AI Agent to Run Daily Operations
By Sarah Williams
Marcus Rush is the founder and team leader of Rush Home, a residential real estate brokerage with eight agents.
Like many growing businesses, his team had already invested in automation. They used tools to connect their CRM with other parts of the business and streamline their workflows.
But there was a problem. Every automation platform came with limits.
If the software didn't offer the right trigger or action, Marcus simply couldn't build the workflow he wanted.
The tools were helping, but they were also defining what was possible.
Marcus wanted something more flexible. He wanted to build a system around the way his brokerage actually worked, rather than changing his processes to fit the limitations of his software.
That'swhere AI changed the equation.
The Solution: Meet "Russ"
Using Claude and Zapier MCP, Marcus built an AI agent he calls Russ.
Russ acts as an AI operations assistant for the brokerage, helping with everything from lead scoring and agent notifications to calendar management.
At the centre of the system is a custom lead-scoring process that runs across a database of more than 11,000 contacts.
Whenever a new interaction comes in, Russ reassesses the lead and considers factors such as:
✔ Whether the person is a buyer or seller
✔ Lead source
✔ Geographic focus
✔ Price point
✔ Response timing
✔ Call duration
The updated score then flows back into the brokerage's CRM.
But Russdoesn'tstop there.
Every morning, it sends each agent a personalised briefing with their highest-priority leads and suggested follow-up approaches based on the information and notes already in the CRM.
Instead of agents starting their day by logging into the CRM and figuring out who they should contact, Russ gives them a prioritised list and tells them where to focus.
From Lead Scoring to Daily Operations
Once Marcus had the lead-scoring system working, he started seeing other possibilities.
If AI could analyse thousands of contacts and prioritise leads, what else could it manage?
Marcus began testing Russ across other parts of the business, including calendar management.
The system can scan email, identify scheduling requirements, manage appointments, and update the calendar through a dedicated workspace account.
He's also exploring other applications, including renter profiling and additional business efficiencies.
The project is ongoing, but the important shift has already happened.
Marcus is no longer limited to the automations that someone else has decided to build into a software platform.
He's building systems around the specific needs of his business.
What Made It Possible?
The key technology behind the build is Zapier MCP, which allows Claude to connect with thousands of applications and take action across them.
Instead of simply generating text or providing suggestions, the AI can interact with the systems Marcus already uses.
That means Claude can work with tools such as the brokerage’s CRM, Slack, and Gmail to move information and actions between them.
For Marcus, this removed a major limitationhe'dexperienced with traditional automation.
Previously, if his CRM didn't have the trigger he needed, the workflow couldn't be built.
With the new approach, he could work from the API documentation and create a custom connection around the way he wanted the system to work.
As he put it:
"We're not limited to a trigger anymore."
That is an important distinction.
The technology is valuable because it gives a business owner much more freedom to design a solution around an actual business problem. Itisn'tvaluable simply because it uses AI.
The Results
The impact for Rush Home is primarily about time, prioritisation, and focus.
With Russ handling theinitialanalysis andorganisationof leads:
11,000+ contacts can be continuously evaluated as new interactions happen.
Agents receive personalised daily briefs instead of manually working through the CRM to determine who to contact.
Lead follow-up becomes more targeted, with suggested approaches based on the information already stored in the CRM.
Calendar management can be delegated to AI, reducing the amount of manual coordination required.
And perhaps most importantly, Marcus can spend more time managing his team and growing the business instead of building and maintaining individual automations.
The Bigger Lesson
Marcus's story isn't really about real estate. And it isn'tabout building an AI agent as complicated as Russ.
Hisstory isabout what happens when business owners can start building technology around the problems they actually have.
For years, if you had an idea for a custom business tool, you generally needed to hire a developer, learn to code yourself, or find a software platform that already did what you needed.
And if none of those options worked?
You lived with the problem.
AI-assisted development is changing that.
Marcus understood his business and knew exactly where the friction was. He didn't need to become a software engineer before he could start experimenting with solutions.
He could describe what he wanted, connect the right systems, test what AI produced, and keep refining it.
That is the fundamental idea behind Vibe Coding.
You start with the problem. You define what a useful solution needs to be.
You direct AI to help build it. Then you test, refine, and improve it.
The technical barrier is getting lower, but the importance of business knowledge is not.
In fact,it'sbecoming more valuable.
The person who understands the problem is often best positioned to decide what to build.
What Could You Build?
Marcus built an AI agent that helps run the daily operations of a real estate brokerage.
Your solution might look completely different.
Maybe you need a dashboard that brings scattered information together.
Maybe there's a repetitive process where your team spends hours managing every week.
Maybe you have an idea for an internal tool that would make your business dramatically easier to run.
Ormaybe you'vealways thought:
"Someone should build a tool that does this."
With AI-assisted development, you may no longer have to wait for someone else.
You can start with the problem you already understand and learn how to direct AI to help you build the solution.
Not sure where to start? Ask us anything or explore where AI could create meaningful opportunities in your business.
Sources:
