Let me be blunt: if you’re not using AI in your real estate business yet, you’re already behind.
This isn’t about hype. It’s about leverage. I’ve been in this industry long enough to know when something’s a gimmick — and when it’s a game-changer. AI is the latter.
I started working in artificial intelligence back in the 90s, long before most people heard about it. I later left the tech world to build and sell a successful real estate brokerage, then went on to found a real estate software company. I’ve been training agents on how to win with tech for over a decade — and right now, there’s nothing with more potential than AI.
In this blog, I’m going to show you exactly how I use AI every day — from generating listing copy that actually sells, to staging properties virtually, to instantly becoming the local expert in neighborhoods I’ve never even set foot in. Whether you’re trying to generate real estate leads, scale your business, or simply reclaim your evenings, this blog will show you how I use AI to handle everything from listing prep to paperwork and negotiation.
This isn’t theory. I’m giving you real examples from my business and showing you how to apply them in yours — step-by-step.
Ready to stop playing catch-up and start using AI like a top producer? Let’s get into it.
How I Use AI to Write Listing Descriptions That Actually Convert
Writing listing descriptions used to be one of the most tedious parts of my workflow. I’d sit down, try to channel my inner copywriter, and spend 30 minutes figuring out how to say “light-filled” in a way that didn’t make me cringe.
Now? I don’t write listing copy anymore. AI does it for me — and better than I ever could.
Let me give you a real example. I had a two-bedroom, two-bath live/work loft in Rincon Hill — unit 402 at 18 Lansing Street. Great property, killer roof deck. I opened up ChatGPT and gave it a basic prompt: “Write a listing description for a 2 bed, 2 bath, 1500 sq ft loft with a one-of-a-kind roof deck.”
Then I did something most agents don’t know you can do: I uploaded photos.
I grabbed the first 10 images — mostly of the roof deck and upstairs area — and let the AI go to work.
What it gave me stopped me cold. It wrote:
“There are two generously sized bedrooms, including a primary suite with direct access to outdoor space.”
I thought, wait — how do you know that?
I hadn’t mentioned anything about a primary suite or outdoor access. So I asked it. And here’s what it told me:
It had analyzed the rooftop photos, noticed a bed visible through the sliding glass doors, and inferred that the room connected to the deck was likely the primary bedroom. It even guessed the bed size (queen or king) and used that to justify its assumption.
Now, was that technically a verified fact? No. But it was a damn good guess — better than what most agents write by hand. And the kicker? It was right.
This is the kind of next-level time-saving and intelligence you get when you stop treating AI like a toy and start using it like a tool.
The takeaway?
Upload your images. Give it context. Let it analyze, infer, and write. Then, of course, double-check everything — AI is powerful, but not perfect.
But once you get the hang of it, you’ll never write listing copy from scratch again.
Let’s face it — writing a standout listing description can eat up your time, especially when you’ve got multiple properties on the market.
Instead of writing from scratch, top agents are uploading property photos into ChatGPT or other large language models and letting the AI do the heavy lifting. These tools can:
- Identify architectural features directly from images
- Infer room types, layout, and amenities
- Suggest compelling language based on your property’s best selling points
Example:
One agent uploaded 10 images of a San Francisco loft. ChatGPT spotted a bed through the window of a rooftop deck and correctly inferred that it was the primary bedroom with outdoor access. It then drafted a description that highlighted this premium feature — all without being explicitly told.
SEO Tip: When you post that AI-generated listing online, optimize the image alt text and include long-tail keywords like “live-work loft with private rooftop deck in Rincon Hill.”
This isn’t just about saving time — it’s about showing off your listing with language that hooks the right buyers. This process is now part of how I write property descriptions that sell, faster and more effectively than ever.
Pair that with strong property presentation techniques, and suddenly your listing stands out before buyers even walk through the door.
How I Became a Local Expert in a Neighborhood I Barely Knew — in Under 15 minutes
Not long ago, a friend reached out to me for advice on selling his home in San Carlos — a suburb in the Bay Area I’ve maybe visited five times in my life. I didn’t know the schools, the subdivisions, or what made one part of town better than another.
Now, normally, I’d start Googling: best schools in San Carlos, top neighborhoods, things like that. But that takes time — bouncing between five different sites, scanning outdated forums and half-baked blogs. You know how it goes. It’s tedious, and frankly, it’s not the best use of my time.
So instead, I let AI do the research for me.
I used a tool called Manus, an AI agent that can surf the web like a human. I gave it one simple job:
“Tell me everything I need to know about San Carlos — neighborhoods, schools, pros and cons.”
What it did blew me away.
It didn’t just spit out a quick list of school names or generic blurbs. It navigated websites, followed internal links, pulled data from real estate sites, school ranking pages, local guides — and compiled all of it into a report I could use right away.
Then I wanted to go deeper. I switched to another tool called Notebook LM, which is great for studying and organizing complex info. I uploaded three local real estate websites into it — agents’ blogs, community pages, and some HOA docs. Here’s what Notebook LM gave me:
- A summarized guide to San Carlos, broken down into subdivisions, pricing trends, school zones, and commute times.
- A clickable mind map — San Carlos in the center, with branches like Neighborhoods, Schools, Transportation, Housing Types.
- A podcast-style audio overview I could listen to while driving — imagine a casual back-and-forth between two voices discussing the vibe, amenities, and lifestyle of the area.
- A custom-written scripted video overview I could’ve dropped into a listing presentation or buyer consult with almost no edits.
And I didn’t have to dig through dozens of articles or cross-check every little stat. AI did the heavy lifting. I just reviewed the results, pulled out what I needed, and used it to guide my friend through pricing, timing, and marketing strategy.
Let me be clear — I didn’t use AI to bluff my way through that conversation. I used it to become an expert fast. And I’ve since done the same thing in neighborhoods I’ve never sold in before.
If you’re expanding your service area or tapping into new real estate niches, this is how you scale your knowledge without burning time you don’t have. Get 500 free credits with Manus.
How I Use AI to Review Offers, Check Paperwork, and Save Hours on Transactions
You know that feeling when a 30-page offer packet hits your inbox at 9:00 p.m. and you’re too wiped to read every line? That used to be me — until I started using AI to break down contracts and summarize the details in minutes.
Let me walk you through a real scenario.
An agent sent me a full offer package — about 30 pages, the usual: purchase agreement, disclosures, CCPA notices, buyer advisories, contingency removal forms, all of it.
I uploaded the entire thing into Manus, the AI agent I mentioned earlier, and gave it a few specific instructions:
- “Create a one-page summary of the offer.”
- “Separate each document into its own file.”
- “Double-check for initials and signatures throughout.”
And it did all of that. Perfectly.
It pulled out every key term — offer price, deposit, down payment, loan structure, closing timeline, seller credits, contingencies, expiration dates. It verified that both buyers had initialed every page of the purchase agreement, flagged that the seller hadn’t signed yet (because, of course, it was a new offer), and even noted that some contingencies were already waived.
Then it organized everything into a neat digital folder with individual PDFs, ready for review or forwarding.
If you’ve ever spent an hour just parsing one of these offers manually — or worse, missed a deadline because you couldn’t get through it fast enough — you know how big this is.
Here’s the kicker: the summary it gave me was so clean I could copy/paste it straight into my CRM, send it to my client, or use it during counter negotiations. No reformatting. No extra typing.
Of course, I double-checked every detail. I don’t hand over legal review to a robot. But what AI gave me was a starting point that saved at least two hours of manual work — and probably some sanity.
If you’re managing multiple offers, tight timelines, or high-volume pipelines, using AI like this will keep you from drowning in paperwork — and let you focus on strategy, not admin.
How I Used AI to Communicate with a $4M Buyer Who Spoke Very Little English
One of the most unexpected ways AI saved a deal for me was with a $4 million buyer from mainland China. He and his family were relocating from LA to the Bay Area and were referred to me by a mutual connection.
There was just one problem: his English was limited. And while I speak Chinese — well enough for casual conversation — it had been years since I used it professionally. I didn’t have the real estate vocabulary dialed in, and I didn’t want to risk a misunderstanding, especially on a deal of this size.
Normally, I’d scramble to brush up. Call a friend, watch a few videos, maybe Google some phrases. But I didn’t have time for that. So I went straight to ChatGPT.
I asked it,
“Can you help me practice speaking Mandarin in a real estate context?”
And it said,
“Of course. What kind of situation do you want to role-play?”
I gave it a scenario — buyer consultation, discussing contingencies, touring homes — and suddenly I had a conversation partner who could answer questions, correct my phrasing, and even suggest alternate ways to explain complex topics like earnest money deposits, seller credits, and close-of-escrow timelines.
What’s more, I could feed it English scripts — my typical talking points — and it would translate them into conversational, culturally appropriate Mandarin. Not robotic. Not word-for-word literal. But natural.
I even used it to write bilingual email templates, WeChat messages, and quick summaries I could send to his wife, who handled most of the communication.
This didn’t just make the process smoother — it built trust. They could tell I was making an effort, and that I could handle the nuance. That kind of rapport is everything in luxury transactions, especially cross-culturally.
Could I have done this without AI? Sure. But it would’ve taken weeks instead of minutes.
This is the kind of leverage most agents don’t even realize is possible yet.
How I Used AI to Screen Tenants, Rank Applications, and Generate Follow-Up Emails — in 10 Minutes
You know what’s worse than having no rental leads? Having six applications to review and zero time to dig through them all. That was me recently when I listed a live/work loft and got a flood of interest within 24 hours. Three full applications came in, each with two applicants, meaning I had six people to evaluate. Rentspree delivered the reports, but they were long — 20+ pages each — and I needed to move fast.
Instead of reading each one line by line, I decided to hand it off to my AI assistant — Manus.
I uploaded the files into Google Drive and gave Manus the link. Then I gave it my instructions:
“For each applicant, check employment length, income, credit score, background check, number of pets, current rent, desired move-in date. Then group each pair into their respective application and summarize everything in a table.”
And that’s exactly what it did.
It read every report, extracted the data, and organized it into a clean, color-coded spreadsheet:
- Employment history
- Combined income
- Credit scores with ratings (e.g. “795 – Excellent”)
- Payment history and any missed payments
- Current address and rent
- Landlord and HR contact info
- Number of pets and personal references
Then it went a step further: it calculated the rent-to-income ratios and flagged which applications had exceptional financials. One pair had a 17.5x rent coverage — far beyond the typical threshold — and the AI highlighted that as “Exceptional.”
Best part? It gave me a ranked recommendation:
- Applicant Pair 1
- Pair 3
- Pair 2
And not just based on income — it considered payment history, employment stability, and application completeness. All I had to do was double-check for accuracy.
I wasn’t done yet.
I asked Manus to write follow-up emails I could send to each HR contact, landlord, and reference. It generated clean, professional templates I could personalize with a few clicks and send out.
What normally would’ve taken me an hour — or cost me a VA — took 10 minutes. And I still had full control over the final decisions.
This is how AI should work for agents: summarize the grunt work, surface what matters, and let you make the call. It’s like having a full-time assistant who never sleeps — and doesn’t need onboarding.
How I Used AI to Create Neighborhood Videos, Curb Appeal Mockups, and Visual Content on Demand
Creating content that looks good has always been a time suck. You either shoot it yourself, hire someone, or dig around the internet for stock photos that sort of resemble what you need.
But not anymore.
I was working on a listing in the San Francisco Peninsula and needed some vertical video clips to use on Instagram and TikTok. I didn’t have any on hand — and I wasn’t about to drive down with a camera and spend my afternoon filming homes.
So I asked Manus:
“Find me vertical videos of homes in the Peninsula — specifically in Atherton, Hillsborough, and San Mateo.”
It started scouring the web — TikTok, Pexels, Shutterstock, and other sources — looking for video clips that matched the criteria. But after a few minutes, it came back and said it couldn’t find exactly what I wanted.
And then it did something unexpected:
It generated photo-realistic vertical videos of homes that looked like they were pulled from Atherton and Hillsborough real estate ads. Not generic animations — I’m talking about clips that looked like they were actually filmed in those neighborhoods.
I gave it feedback:
“This looks too Victorian — make it more Peninsula modern.”
“Don’t use North Bay or Tiburon-style homes. Stick to these zip codes.”
It listened. It refined. It delivered.
The result? A series of short, branded, highly shareable video clips tailored to exactly the market I was targeting — all generated by AI, no filming required.
I also needed curb appeal suggestions for a seller in San Carlos. I didn’t have time to send it to a stager or wait for an architect. So I snapped a quick photo of the front of the house and uploaded it to ChatGPT and Manus with this prompt:
“Give me cost-effective improvements to increase curb appeal and show me what it would look like.”
Here’s what it suggested:
- Add layered planting along the walkway
- Repaint the garage door in a modern gray
- Highlight the entryway with a more colorful front door
- Remove the outdated iron railing
- Power wash or refinish the driveway
- Trim the tree and replace bare patches with fresh mulch
It even rendered two versions of the photo with the upgrades applied — one from Manus, one from ChatGPT. Were they perfect? No. Did they give me a solid visual to talk through with my seller? Absolutely.
One even picked up on the misaligned front path and suggested straightening it — which, coincidentally, the sellers had already been considering. I walked into that listing meeting with mockups, design suggestions, and a vision. No designer. No stager. No delay.
This is what modern listing prep looks like.
You bring the expertise. Let AI handle the production value.
These visuals aren’t just time-savers — they boost engagement across every channel, from Zillow to real estate social media marketing.
If you’re building out a content pipeline, AI-powered real estate advertising ideas like this will absolutely change how you market.
How I Pulled a CMA, Gave Seller Advice, and Prepped for a Listing Meeting — All With AI
Let’s go back to that San Carlos listing.
After I used AI to learn the neighborhood, I needed to pull recent comps to prepare for a meeting with the seller. I didn’t have access to the local MLS at that moment — and honestly, I just needed a ballpark valuation fast.
So I asked Manus to gather recent sales within 0.25 miles of the property at 74 Cedar.
At first, it gave me a pretty wide price range. So I did what I always do with AI — I pushed it harder.
“Narrow it down. Focus on homes with 4 bedrooms. Match the square footage. Ignore outdated sales.”
The AI adjusted its criteria, cross-referenced sales from Redfin and Zillow, and came back with a refined CMA:
Properties within $50K–$100K of each other, sold recently, with similar beds, baths, and condition. Not perfect — but solid enough to frame my strategy and walk into the listing meeting with confidence.
Then the seller asked me,
I had my own ideas — I always do — but I decided to test the AI again. I uploaded a photo of the exterior and asked:
“What are the most cost-effective improvements we can make to increase curb appeal and salability?”
It suggested:
- Repainting the front door for contrast
- Adding greenery and flower beds (using the existing mulch and rocks)
- Updating exterior hardware
- Replacing or removing the iron railing
- Cleaning or redoing the cracked driveway
- Straightening the crooked front path
It even generated a visual mockup of the changes — and it was dead-on. When I met the sellers, they mentioned straightening the path and updating the driveway before I even brought it up. AI gave me data-driven validation for my recommendations and helped me present like I’d already done a walkthrough.
If you’ve ever needed to prep fast, walk into a listing cold, or just want to look like the most prepared agent in the room — this is how you do it.
Your Next Steps: How to Start Using AI in Your Real Estate Business This Week
Now, AI tools like ChatGPT and Jasper can generate compelling, SEO‑friendly property descriptions in seconds.
Here’s what I’d do if I were starting from scratch:
- Pick a general AI assistant and stick with it.
Start with ChatGPT (paid) or Manus. Both can handle 80% of what you need.
- Use AI on your next listing.
Upload property photos, generate listing copy, try virtual staging, and test photo enhancements.
- Practice using it for paperwork.
Summarize a contract. Review a PDF disclosure package. Use it to pull out deadlines and terms.
- Train with it.
Role-play with fake buyers, sellers, objection scenarios — even in other languages if you need.
- Research like a pro.
Use tools like Notebook LM to prep for unfamiliar neighborhoods or buyers who want detailed local knowledge.
- Get lazy on purpose.
That’s right. Start pushing tasks to AI and see how far it can go. You’ll be surprised how much admin work you no longer need to do manually.
The free versions are fine, but the real power — and time savings — come with premium features. It’s not a cost, it’s leverage.
- Pay for the tools.
The free versions are fine, but the real power — and time savings — come with premium features. It’s not a cost, it’s leverage.
