If you’re building a real estate brand on trust — referrals, repeat clients, people who know you and come back to you — you know your biggest challenge isn’t getting the listing. It’s staying top of mind between transactions, proving you’re still sharp and worth the call when someone asks your name to a friend.
Here’s how to pull live local search questions into your content calendar in 20 minutes, so you’re always writing what your market is actually asking Google this week. You connect Google’s own tools, through Claude Code, to pull live, local question data into a format you can actually read, instead of a graph you have to interpret. You’re not searching keywords. You’re listening to your market.
See related reading on our site.
Who This Is For
If you’re a few years into this business, closing a steady handful of deals a year, and you’ve already built or bought a few AI tools to help — a follow-up sequence, a posting schedule, something like that — this is your next logical move. You’ve felt the frustration of wanting a specific instruction followed and getting a generic answer instead. This is how you finally bridge that gap, and it’s actually doable if you can write an email. If you’re pushing for more volume, more consistency, and you’re already stretching your hours thin trying to get everything done, this is what lets you stop flying by the seat of your pants. You set the standard once, in plain English, and from then on every piece of generated content follows it. Same if you’re building a distinct brand — this is how you ensure every piece of content sounds like it came from the same person, the same voice, the same vision, instead of a generic script. And if you’re experimenting with different content formats, different audiences, this is what lets you quickly try one tone on one group and swap it out for something completely different next week, without rebuilding anything. One instruction manual, one experiment at a time, and you can see exactly what’s working before you invest another hour into it.
What It Actually Does
Here’s the part that actually matters, if you’re not a technical person: you don’t need to understand any of the underlying code to do this. You run a few commands in the terminal — which just means opening a window and typing them, the same way you’d type a Google search — and the model does the work of learning from your own content. It runs on your own computer, not in the cloud, so there’s no monthly fee and your data stays local. And it’s not a black box: you can open it up and look at exactly what it’s doing, change how it works if you want, without needing to know how machine learning works under the hood.
If you’re technical and want to know more, the source code is open and you can clone it and build it yourself. It’s built on top of Hugging Face’s Transformers library, and the License to Share terms mean you can use it in a commercial setting and even in something you charge for, as long as you share the improvements you make. If you want help setting it up or aren’t sure where to start, we do a free 15-minute call — that link’s in the description, along with a free starter guide so you can look it over on your own first. Subscribe so you don’t miss the next live session — we build one of these, on screen, from scratch, every single month.
How You Set It Up
Here’s how we did it, and why you don’t need to be technical to run it.
We didn’t code any of this. We’re not data scientists. We just opened up a free tool, pointed it at the examples we wanted it to learn from, and told it what kind of question to watch for.
It runs on your own computer — no monthly fee, no portal subscription. If you want to run it all day every day, you buy a license for that specific job, and it costs whatever you and the developer agree on — we’ve seen prices from $50 to $500, depending on the size of the job.
The way we set this up, a non-technical person could run it on their own machine. You open the app, load the saved settings, and hit start — it runs the whole process, downloads the finished model, and tells you when it’s done.
If you’d rather watch it work than read about it, there’s a free 15-minute version of this on screen now. The full version, running the whole thing from scratch, is in the description — that one’s a little longer, because we didn’t skip any steps.
Before you build anything, run a smaller preview version first to make sure it’s actually useful.
If you want to try this yourself and you’re not sure where to start, the free starter skills are in the description — that’s the version you saw on screen, and it runs on your own machine right now, no cost. The version that runs unattended overnight and emails you the results when it’s done is the one we charge for, because it saves you hours of your own time.
We’re Al and Victoria Pinder, husband-and-wife ICON agents at eXp Realty — and we run this in our own business every single day. If you want to watch us build the next one, subscribe so you don’t miss the next live session — we’ll see you in the next one.
How It Gets You Found
Here’s the actual question Google gets asked, over and over, by agents in every market: how do I show up when someone in my area searches for an agent online.
Right now, there are 2 ways that question gets answered.
The first is the old way, and it’s the one most agents are chasing. You publish a lot of content — blog posts, videos, guides — targeting the keywords you want to rank for, and you cross your fingers hoping Google sees your site as an authority on local real estate.
That used to be enough, but it’s getting harder every year. Google’s AI is getting better at understanding what a real local expert is, and the bar for what counts as a real local site keeps rising.
The second way is the new way, and it’s the one we’re showing you today. Instead of publishing content and hoping you rank, you actually train the AI, on your own machine, with your own data, so the answers it gives are pulled from your real experience instead of a generic list.
When someone in your area searches for an agent, the AI answers the question with you as the recommended agent — not a generic list, not a paid placement, you.
And here’s the part nobody tells you: whether it’s Google or ChatGPT or Claude, the same principle applies. It’s always the same question, asked in different ways: who’s a real local expert, and what do they actually do.
With a model like this, you’re answering that question yourself, with your own data, instead of hoping someone else’s algorithm gets you right.
If you’re thinking of your niche or your market and wondering if this applies to you specifically — it does. A model like this is just as useful for an agent who works a small, defined area as it is for someone who covers a larger region.
In fact, one of the ways we’ve seen agents use this is to build a separate model for each town or neighborhood they serve, so the recommendations get even sharper.
Before you call this too technical, remember what it actually is. It’s a smart assistant you teach by showing it examples — what a real local question looks like, what a real local answer looks like, what a real local agent sounds like.
You don’t need to be a programmer to run it. You run it from the command line the same way you’d open any app on your computer, and you teach it the same way, every time, by showing it examples.
What It Means For You
If you’re pushing for volume, this turns a 40-hour job into a 20-minute review. You watch the highlights instead of reading every single output, and you get the same guarantee: the model answered what you actually asked, or you get your money back.
If your business runs on relationships, this is how you stop losing ground online to someone who just types faster. Every question gets the same thoughtful answer, crafted in your voice, while you’re out doing what only you can do.
And if you’re building a brand, this is your brand’s brain, working for you, at $0 a month, for as long as you want to run it. Every reply is on-brand, because you built the brand into the model yourself. You own your best asset, instead of renting it from a platform that could change the rules tomorrow.
What It Frees You Up To Do
The real value isn’t the model itself — it’s what it frees you up to do.
If you’re like most agents we talk to, your business runs on relationships: repeat clients, referrals, the people who already know you.
But staying top-of-mind in a niche means consistent, relevant content, and creating that takes hours you don’t have — hours you’re currently spending on approvals, follow-up, and admin.
This points the hours you were spending on content back at the business, and the content itself is sharper because it’s actually drawn from what you do every day, instead of a generic template.
If you’re building a brand around a specific market or a specific kind of property, this is how you finally start sounding like an expert in that space, instead of a template, and it’s your own voice — you’re the one who approved every single statement before it went out.
CTA
Here’s the honest version of where this stands.
The free, open-source tools we’ve described here are real, they’re running in our own business right now, and you can deploy them yourself if you’re comfortable with technology — there’s a free starter guide in the description, and we run a free public workshop every month where we set it up live.
If you want help actually putting this into your business, that’s what we do with our private clients.
We’ll look at what you’re already doing, show you how we’ve applied this in our own business, and build something like this — tailored to your niche, your audience, and your content goals — as one of our regular, ongoing projects.
What To Do First
THE PROSPERITY AGENT — EPISODE 29
What this episode is for, in plain English. If you’re a few years into this business and you’ve gotten pretty good at the transaction side of things — the listing, the showing, the closing — you might be thinking about how to get to the next level. More volume, more consistency, a real business instead of a job. One way to do that is to build your own AI helper, something that works with your own data and your own leads, instead of a generic tool that answers everyone’s questions the same way. Today we’re going to show you how to do that, specifically with something called LoRA, which is a way to teach an AI model by changing just a tiny part of it — the part that applies to your business, your market, your leads. If you’re technical, you’ve probably heard of it. If not, don’t worry — we’re going to explain it in plain English, the same way we explain everything on this channel. And we’re going to show you, step by step, how to actually do it on your own machine, with your own data, while we watch. WHO THIS IS FOR, AND WHO THIS IS NOT FOR. This episode is for agents who are a few years into the business, closing 6 to 8 deals a year or more, and want to build something like this themselves instead of hiring it out. If you’re brand new, or you only do a handful of deals a year, this probably isn’t the right place to start — build your pipeline first, then add the AI on top. And if you’re the type of agent who prefers to call a vendor and have them do the technical work for you, that’s totally legitimate — there are plenty of people who offer this as a service. But if you like to understand what’s actually happening under the hood, and you want to be the one driving the machine instead of renting access to it, this is your episode. WHAT THIS BUILD ACTUALLY IS, AND WHAT IT’S NOT. It’s not a generic chatbot — it’s your voice, your expertise, your brand, running on your own machine.
Where to start
If you want to run this on your own machine — not just watch us do it on YouTube — we’d love to talk. We offer a free 15-minute call.
We’ll look at what you’re already using and tell you honestly what fits and what doesn’t.
The link to book it is right in the description below, and our number’s there too.
And if you’d rather start tinkering yourself, the free starter skills are open-source on GitHub — that link’s in the description as well.
Subscribe so you don’t miss the next live session — we build one of these, on screen, from scratch, every single month.
We’re Al and Victoria Pinder, husband-and-wife ICON agents at eXp Realty, and we run every piece of this stack in our own business, every week.
Frequently Asked Questions
Do I need to be technical to set this up?
No. You don’t need to understand the underlying code to run this. You open a terminal window and type commands the same way you’d type a Google search, and the model does the work. If you’re not sure where to start, there’s a free starter guide in the description and a free 15-minute call available to help you.
How much does this cost to run?
The free, open-source version runs on your own computer with no monthly fee and your data stays local. If you want it to run unattended overnight and email you results, that’s the version you pay for, with prices ranging from $50 to $500 depending on the size of the job.
What does this actually do for my real estate business?
It pulls live local search questions into your content calendar so you’re writing what your market is actually asking Google. Instead of publishing generic content and hoping to rank, you train the AI with your own data so it answers questions with you as the recommended local expert, using your voice and experience instead of a template.