SRAR Member Resources · Thursday, August 20, 2026

AI + HI: The Modern Realtor

Everything from the talk at the Southland Regional Association of REALTORS. Downloads, workflows, blueprints, free and with no email wall. I am a licensed California REALTOR who builds AI systems for other agents. Same license, same liability, same market as you.

Book Your Free 15-Minute AI Audit Get the Downloads
California REALTOR since 1998 20 Yrs LAPD 1,000+ agents trained in AI A licensed agent who builds the AI DRE #01238257

Watch the Full Presentation

The complete talk as delivered in the room, 62 minutes, on demand. Chapter markers are on the video so you can jump straight to any system before you build it.

Presentation Downloads

Scanned the QR at your table? You are in the right place. Every asset from the talk, free, no email wall.

PDF · 2 Pages

The SRAR AI Field Guide

The leave-behind. Six rules, three jobs for AI in your practice, the four moves, and the 90-day plan on two printable pages.

Download the Field Guide ↓
PDF · Action Guide

Systems Over Prompts: Action Guide

The vending machine trap versus the digital employee, plus the high-yield automations you can start building Monday.

Download the Action Guide ↓
PNG · Mind Map

The AI Systems Mind Map

The whole architecture on one screen. Context, memory, tools, and skills, mapped the way the talk laid them out.

Download the Mind Map ↓
PNG · Infographic

Real Estate AI: Evolution Comparison

Where the reactive agent loses and the systems-driven agent wins, side by side.

Download the Infographic ↓
PDF · 4-Page Workbook

The AI-Powered Agent Implementation Playbook

The full SOP workbook: lead clustering playbook, AI employee onboarding script, outcome-driven prompt grid, and the agentic boardroom matrix.

Download the Playbook ↓
PDF · 15 Slides

The Presentation Slides

Every slide from the talk, in order, with the frameworks intact. Speaker notes included.

Download the Slides ↓
PDF · Toolkit Part 1

The Context File Interview

Do not write your context file, talk it. Four prompts that make any paid AI interview you by voice, write the file, plan your 90 days, and keep itself updated weekly.

Download the Interview ↓
PDF · Toolkit Part 2

The Straight Talk Rules

Every AI ships tuned to flatter you, and that tuning costs you money. Ten paste-in rules that turn the machine from a fan into the sharpest advisor you ever hired.

Download the Rules ↓
PDF · Toolkit Part 3

The Compliance Guardrails

The fair housing, client data, and verification rules that keep your license attached to your name, written as hard constraints you paste into the machine once.

Download the Guardrails ↓
PDF · Toolkit Part 4

The Private AI Workup

Your own AI on hardware you own, inside your own walls. The one setup where contracts, client files, and protected documents can finally go in, because nothing ever leaves the building.

Download the Workup ↓
PPTX · Editable

The Slide Deck, Editable

The PowerPoint file itself. Speaker notes are embedded on every slide. Use it, adapt it, present it at your own office meeting.

Download the PowerPoint ↓

The 11 O'Clock Problem

The reason any of this matters.

Last Tuesday night, 11 o'clock, a homeowner four miles from my desk sat up in bed, picked up her phone, and typed one sentence into a machine.

Who are the best real estate agents in Santa Clarita?

She was not looking for a house. She was not filling out a form on Zillow. She was deciding whether to sell the place she raised her kids in, and she was doing it at 11 at night, because that is when people make the decisions that change their life. Before she asked a single agent, she asked a machine which one of us to trust. The machine answered. It gave her three names.

You were not in the room. You did not get a notification. There was no lead. There is no record. You will never know it happened.

You were on the list too. You were just faded out. Not because the machine dislikes you, and not because you are bad at this. It could not find you.

So you are not behind on AI. You are invisible. Those are two completely different problems. Behind means catching up, and catching up is exhausting and never ends. Invisible means the machine standing between you and your next seller does not know you are there. One of those you fix by working harder for the rest of your career. The other one you fix by building one system, one time. And to be clear about what this page is: we are not building it for you while you read. You walk out of here with the knowledge of how to build the proper system, and everything below is that knowledge.

What Is This AI Thing Anyway?

Most of the fear in this industry comes from arguing about something nobody has defined. So define it.

It is not software. Software follows rules a programmer wrote, and it runs the same way every time, forever. Your MLS portal is software. Your transaction platform is software.

AI is different. You talk to it in plain English, it reads what you give it, and it writes back. That is really it. It drafts, it sorts, it summarizes, it follows up, and it never gets tired.

It was never AI versus HI. Change one character: it is a plus sign. The machine brings speed and memory. You bring the license, the judgment, and twenty years of handshakes in this valley. Neither one closes a deal alone.

And here is the part nobody selling you a course will say out loud: a lot of what is being promised to agents right now, the AI that runs your whole business while you sleep, the robot that replaces your assistant on day one, is either fantasy or not there yet. What is real today is simpler and more useful: drafting, sorting, summarizing, and follow-up. That is plenty, and it is what everything on this page is built from.

The Receipts: What 59 Videos and 1.46 Million Views Teach, and What They Skip

Before the talk I pulled the most-watched YouTube videos teaching real estate agents AI: 59 videos, 1.46 million combined views, 21.5 hours, 253,576 words of transcript, analyzed line by line. Two lists came out.

List One

What they all teach: cold outreach and scripts (46 of 59), email and follow-up (45), automation (40), prompting (34), social media and video (31 each). None of it is wrong. All of it is everywhere, which means it makes you average, faster. And watch the endings: the lesson is free, the dream costs money. If autopilot lead machines existed, the people selling them would use them instead.

List Two

What 1.5 million views never told you: fair housing gets mentioned in 2 of 59 videos. Where your client data goes, 1 of 59. Verify before you send, 8 of 59. Your license, your broker, compliance, 4 of 59. Fifty-one of fifty-nine never tell you to check a single thing before it reaches a client. Caution does not get views. But you do not work in views. You work in a licensed profession, and that gap is the whole game.

What the Talk Covered

Three pillars. Eight systems. Zero shiny objects.

PILLAR 01

Systems

Stop treating AI like a vending machine. Build a Chief of Staff with context files, memory, tools, and skills that carry into every conversation.

PILLAR 02

Speed

A trained officer under stress hits 30 percent at seven feet. A master framer drives a four-inch nail into redwood in one or two swings, every time. Same hammer. Replace hope with repeatable mechanics.

PILLAR 03

Human Leverage

Let the machine triage the inbox and collate the comps. You sit at the kitchen table, shake hands, and protect your clients.

01

The Lead Clustering Playbook

Run 10,000 dusty CRM contacts through a secure pipeline, cluster by Santa Clarita tract codes, cross-reference live MLS activity, and surface the streets with low inventory, high turnover, and 7-to-12-year equity-rich owners. Ten minutes of prep, and you arrive as the data authority for their block.

02

AI Engine Optimization and the Short List

Half of American adults now use AI chatbots and 60 percent read the AI summary sitting on top of their search results. The answer names a handful of people, not a page of ten blue links. Perfect entity consistency across your name, brokerage, DRE number, phone, Zillow, and Yelp decides whether you are one of the names or invisible.

03

Your AI Employee: Context, Memory, Tools, Skills

Interview yourself in voice mode, let the model transcribe and write its own onboarding files, then connect calendar, email, and CRM. No code required.

04

Outcome-Driven Prompting

Level 1: write me an email. Level 2: here is my goal, my database, my comps, and my brand file. Build the plan. Declare the objective first and the model aligns its reasoning behind your targets.

05

The Shiny Object Blocker

Instruct your AI to push back hard and tell you no until your current system is 100 percent live. Then spend the saved hours on five high-value human connections a day.

06

The Coordinated Agentic Boardroom

One Orchestrator you talk to. Sub-agents for transactions, marketing, and inbound voice, each on least-privilege access with human checkpoints before anything reaches a client.

The 30 Percent Problem

This is a mechanics problem, not a talent problem. Here is how I know.

In law enforcement we trained constantly. Range time, qualification drills, over and over, for years.

When officers get into an actual shooting, they hit what they are aiming at roughly 30 percent of the time. Seven times out of ten, at less than seven feet, they miss. Seven feet is closer than you are standing to somebody at an open house.

That is not a training problem. These are trained people. Under stress your fine motor skills degrade. Tunnel vision sets in. You get audio exclusion and you lose your hearing. Everything you rehearsed goes out the window and you are operating on hope.

The ones who hold up are not braver and they do not have different DNA. They have repeatable mechanics that keep running when the thinking part of the brain checks out.

Now hand me a hammer and a box of nails.

I will miss more than half my swings. I bend the nail over, prop it back up, hit my thumb, and end up at urgent care. Hand that same hammer to a framer and he drives a four-inch nail into a redwood four-by-four in one or two hits. He has driven ten thousand nails. He is still swinging freehand though. Every nail from scratch, relying on his eye, his arm, and whatever kind of day he is having.

Now give that same man a nail gun, a jig that sets the angle, and a depth stop that decides exactly how far it sinks. Same man, same nail, same wall, and every single one goes in identical.

He did not get better. His system got better.

You do not need to get better at real estate. You have the license, you have the market knowledge, you have the thing you do with people. Some of you have been doing this longer than I have. You do not need more talent. You need a jig, a depth stop, and power behind it.

So where are you still swinging freehand?

The listing appointmentNo structure. You walk in hoping the rapport carries it.
The CRM nobody has opened in three weeksThe database is the single most valuable asset you own and most of us treat it like a junk drawer.
Praying the phone ringsHope has never been a lead source and it never will be.
Follow-up that restarts from zeroEvery single day, from nothing, off memory.

What the Data Actually Says About Getting Found

You have been told SEO is dead and that AEO replaced it. That is wrong, and if you act on it you will hurt your business. Here are the real numbers, with sources, so you can check them yourself.

15 → 8

Across 68,879 Google searches, people clicked a traditional result 15 percent of the time when there was no AI summary on the page. With one, 8 percent. The click got cut roughly in half. Clicks on links inside the summary itself: 1 percent.

Pew Research Center, July 2025
34.5 → 58

Across 300,000 keywords, the number one organic result lost about 34.5 percent of its clicks when an AI Overview sat on top of it. Ahrefs re-ran the study on December 2025 data and measured 58 percent.

Ahrefs, 300,000 keywords
~0.2

Percent of all web traffic currently arriving as an AI referral, measured between 0.15 and 0.25 percent. The bike path is real. It is still a bike path.

Similarweb, 2026
49 / 60

Percent of US adults who now use AI chatbots, up from 33 percent in 2024, and the percent who read the AI summary sitting in their search results. Adoption is the story. Referral traffic is not, yet.

Pew Research Center, February 2026

So ranking number one still matters. It just pays about half of what it used to.

Anybody telling you to abandon SEO for AEO is telling you to abandon the highway for the bike path. You build both. SEO gets you into the index. AEO gets you named in the answer. Same foundation, two jobs.

The Cheapest Fix in the Entire Talk

This one costs you an afternoon and zero dollars. If you were half listening, come back for this.

Your legal name, spelled identically everywhere. Your brokerage and your DRE number matching on every profile. One phone number, not the old one still sitting on three old listings. Same headshot, same bio language, same service area. Google, Yelp, Zillow, your website, your social profiles. There are 80-some sites out there pulling your information for real estate.

When the machine sees conflicting details it cannot resolve who you are. It is a computer. Connor T. MacIvor, Connor MacIvor, and Connor With Honor read as three different people with three different phone numbers.

So it moves on to somebody it can verify. Do this one this week.

Build an AI Employee, Not a Vending Machine

Four parts. If you remember nothing else from the talk, remember these four words.

PART 01

Context

The onboarding document. Your biography, what you specialize in, what you refuse to do, your brokerage structure and splits, the neighborhoods you know cold, and your Fair Housing parameters written down explicitly. You would spend a week teaching a new assistant this. You do it once here, and it never quits, never forgets, and never goes to work for a competitor.

PART 02

Memory

Persistent memory is the thing the free tiers do not give you, and the entire system depends on it. This is the reason step one is a paid reasoning workspace.

PART 03

Tools

The hands. Calendar, inbox, CRM, whatever you already pay for, so it can see your actual week. Two rules keep you licensed here: least privilege, and a human checkpoint before anything reaches a client.

PART 04

Skills

Your procedures, written down once, in order, the way you would explain them to a brand new agent. Save each one under a name and invoke it instead of re-explaining. The machine does not have bad Tuesdays.

Nobody wants to sit down and write a document about themselves. That is why everybody skips this step, and skipping it is exactly why their output sounds like a brochure.

So do not write it. Open voice mode on your phone, hit record, and interview yourself for ten minutes. Let the machine transcribe it and organize it into the file, then tell it to ask you more questions until it has what it needs. Everybody will talk about themselves for ten minutes. Almost nobody will write the page. And you do not have to invent the questions: The Context File Interview in the downloads above is the exact prompt series that makes the machine run the interview for you, write the file, and build your 90-day plan from it.

Then teach it how you actually talk, not how you think you should write. Formal and polished is not your voice, that is your nervous voice. Your real voice is the one you use at the kitchen table at 7 at night when a seller asks whether they should take the offer. It is direct, it is short, and it never uses the word furthermore. Paste in emails that actually got a reply, not your best writing. Then name the words that make you sound like a brochure and forbid them. Mine are stunning, nestled, boasts, and any sentence that opens with welcome home.

The test: if your spouse can read it and not know a machine wrote it, you are done. If they catch it in the first sentence, keep going. It takes about three rounds. Anybody telling you it is one shot has not actually done it.

Practice Partners: Four Prompts That Make You Better, Not Lazier

In the talk I debunk prompting. A prompt is the vending machine, the bottom rung, and most of what gets sold as prompt packs is commodity advice. But it is where you start, and these four are different on purpose: they do not write anything for you. They turn the machine into the hardest person in the room, and it coaches you between rounds.

How they work: paste one into a paid AI workspace and answer out loud or by typing. The machine stays in character, pushes back, and after every answer it steps out of character and reflects back at you: what was weak, what a sharper answer sounds like, and where you brushed a compliance line. Run one before every real appointment. Ten minutes of reps beats an hour of hoping.

1. The Critical Seller

Run this before every listing appointment. The machine plays the homeowner who has already interviewed two other agents.

You are playing a homeowner in Santa Clarita, California who is interviewing me, a real estate agent, for the job of selling your house. You are smart, you are skeptical, and you have already interviewed two other agents. You are not rude, but you push.

Ask me one question at a time, and cover at least: my pricing strategy and how I defend it, my marketing plan in specifics, how often I communicate and by what channel, what happens if the home sits, what I do that a discount option does not, and how I handle multiple offers. Challenge vague answers. If I dodge, ask the same question again, harder.

After each of my answers, break character and give me a COACH'S NOTE: score the answer 1 to 10, name exactly what was weak, show me what a sharper answer sounds like in my own words, and flag anything that comes near a fair housing or compliance line. Then go back into character and continue.

After ten questions, end the interview and give me a debrief: my three weakest answers rewritten stronger, the one habit to fix before a real appointment, and your hire or no-hire decision with the reason. Do not go easy on me. I get better when you are critical.

This practice drill is from the AI + HI system by Connor T. MacIvor, ConnorWithHonor, REALTOR since 1998, CalDRE #01238257, Sync Brokerage, Inc. DRE #02031490.

2. The Critical Buyer

The buyer consultation is where representation gets won or lost. The machine plays a careful buyer who does their homework.

You are playing a careful, well-researched home buyer in Santa Clarita, California who is interviewing me, a real estate agent, before agreeing to work with me. You read everything, you have talked to friends who just bought, and you have hard questions.

Ask me one question at a time, and cover at least: how my representation works and how I am paid, explained in plain English; how I win in a multiple-offer situation without overpaying; how I handle inspections and repair negotiations; what I know about specific Santa Clarita neighborhoods and what I check that a portal cannot show; and what happens if I find them a home the first weekend versus month six. Push on anything that sounds rehearsed.

After each of my answers, break character and give me a COACH'S NOTE: score it 1 to 10, name what was weak, show me a sharper version in my own words, and flag anything vague about representation, compensation, or fair housing, because vague is where agents get in trouble.

After ten questions, end it and give me a debrief: my three weakest answers rewritten stronger, the question I most need a better answer for, and your decision on whether you would sign with me. Be critical. That is the job.

This practice drill is from the AI + HI system by Connor T. MacIvor, ConnorWithHonor, REALTOR since 1998, CalDRE #01238257, Sync Brokerage, Inc. DRE #02031490.

3. The Lender Translator

Agents lose deals in the gap between agent-speak and lender-speak. The machine plays a senior lender who trains you to close that gap.

You are playing a senior residential mortgage lender with twenty years of experience, and your job is to train me, a real estate agent, to speak your language fluently so I can protect my clients' deals.

Work in three modes and rotate between them. Quiz mode: test me on the difference between rate and APR, points, DTI, LTV, pre-qualification versus pre-approval versus underwritten approval, rate locks and lock extensions, seller credits versus price reductions, and appraisal gaps. Roleplay mode: play the lender on a live file and make me coordinate a wobbling deal with you, a nervous buyer, and a deadline. Translation mode: I say something the way an agent says it, and you show me how a lender hears it and what I should have said instead.

Correct my terminology immediately and never let an imprecise term slide. After each exchange, give me a COACH'S NOTE: what I got right, what a lender would flag, and the exact phrasing that would have made me sound like I have done this a thousand times. Do not give me consumer financial advice and do not quote today's rates; this is vocabulary and coordination training for a licensed professional.

After a full session, hand me a one-page cheat sheet of the terms I fumbled, with plain-English definitions in my voice.

This practice drill is from the AI + HI system by Connor T. MacIvor, ConnorWithHonor, REALTOR since 1998, CalDRE #01238257, Sync Brokerage, Inc. DRE #02031490.

4. The Property Management Interview

If you handle rentals or want the referral, this owner will find every hole in your pitch before a real one does.

You are playing a Santa Clarita rental property owner interviewing me to manage your property. You have one good rental, you have been burned by a bad tenant before, and you are deciding between me and a big property management company.

Ask me one question at a time, and cover at least: how I screen tenants and how my screening stays fair housing compliant; how fast maintenance calls get answered and by whom; how rent collection and owner statements work; how deposits and funds are held and accounted for; how I market a vacancy and what my average days-vacant looks like; how inspections are scheduled; and what happens, step by step, if a tenant stops paying. Push hard on the burned-before fear.

After each of my answers, break character and give me a COACH'S NOTE: score it 1 to 10, name what was weak, show me a sharper version in my own words, and flag anything in my screening or advertising answers that brushes a fair housing line, because that is where this business ends careers.

After ten questions, end it and give me a debrief: my three weakest answers rewritten stronger, the compliance gap I most need to close, and whether you would hand me the keys. Be the owner I hope I never get, so the real ones feel easy.

This practice drill is from the AI + HI system by Connor T. MacIvor, ConnorWithHonor, REALTOR since 1998, CalDRE #01238257, Sync Brokerage, Inc. DRE #02031490.

One rule carries over from everything else on this page: these are practice partners, not ghostwriters. When you move from practicing to producing, the rule stands. AI drafts. You verify. You sign.

Your Actual First Week

Do not try to build all of it tonight. A concept does not survive the drive home. A calendar does.

Monday
One paid workspace

ChatGPT, Claude, Gemini, Copilot, or Grok. One. Nothing else. Do not browse for tools.

Tuesday
Ten minutes of voice notes

Interview yourself about your business. Transcribe it and save it as your context file.

Wednesday
One skill written down

Your open house follow-up is the easiest place to start. Every step, in order.

Thursday
Connect one tool

Calendar, inbox, or CRM. One job to start: categorize new leads and fire the follow-up.

Friday
Run it once and fix it

That is a complete first week, and it beats what 90 percent of people who watch this will actually do.

Seven Mistakes That Cost Agents the Moat

Fast, because you will recognize most of these.

The guardrail that makes all of it survive contact with your actual week:

Do not start anything new until your current system is live and working. One system at a time. Finish before you start. Then spend the time you win back on five real human conversations a day.

The Rules That Keep You Licensed

A vendor will not tell you any of this. A vendor sells. I am telling you because you are licensed and so am I, and that is the difference.

Protected documents never go into any AI tool, free or paidPre-approval letters, executed contracts, lockbox codes, and anything covered by financial, medical, or credit privacy law. Training off and incognito mode do not make it local: the data still leaves your machine, still gets processed on someone else's servers, and still sits in their logs. Settings change what they keep, not where it goes. For everyday client work, use a paid workspace with training turned off, send the minimum the job needs, and strip names and account numbers the task does not require. If it must stay private, it runs on a local model on your own hardware or it stays out. If you write down one thing from this entire page, write down that one.
Fair Housing applies to what the machine writesAsk for a listing ad with no guardrails and it will hand you "charming starter home in a safe, exclusive neighborhood, perfect for young families, walking distance to St. Mary's." Three violations in one sentence, written in your voice, under your name, in two seconds. It covers AI-generated marketing copy, ad targeting, and anything that screens people. Disparate impact means intent does not save you. The liability lands on you and your broker, not the software vendor. Write your Fair Housing parameters into the context file explicitly, because the model follows what you said and will never know what you meant.
Least privilege on every connectionYour marketing assistant does not need access to your bank account. Scope every agent narrowly.
A human checkpoint before anything reaches a clientNo exceptions, no matter how good the draft looks.
If you self-host, sandboxing ships off by defaultTurn it on before you connect a single account. These harnesses have had real vulnerabilities, including a zero-click takeover bug and malicious plugins found on an official hub. You hold client financial data and you carry record-keeping obligations. Nothing runs unattended.

AI only drafts. You verify. You sign. You ship.
Every time. No exceptions. Your license is on that document. The model's is not.

The Private AI Option: Your Own Model, Inside Your Own Walls

Every rule above says the protected documents never go into any cloud AI tool. So where can they go? Into a machine you own.

I installed my own personal large language model in my residence, on compute I purchased and set up myself. It runs every day, and nothing it reads ever leaves the building.

That is the one configuration where the privacy warning disappears. No account, no provider, no logs on anyone else's servers, no wire for the data to cross. Executed contracts, pre-approval letters, client files, an attorney's case files, a brokerage's entire transaction history: a private system can hold all of it and answer questions against all of it, because it never goes anywhere.

This is a service I provide as the advisor and builder. For agents and teams who want their client files searchable without a privacy question. For brokerages that want a firm-wide AI trained on their own policies, forms, and market history, where the firm's data stays the firm's. For attorneys who need privileged files to stay privileged. I help you spec and order the right components the first time, install the model and the private document index on site, train it on your business, and hand your staff the keys with written procedures.

Straight talk on the trade: local models trail the big cloud models in raw intelligence. Their jobs are reading, sorting, summarizing, and drafting from your own documents, which is most of the daily work, and exactly the work you cannot send out. The smart shape is both: cloud AI for public-facing work, the private system for everything the cloud must never see. And the rule survives even here: AI drafts, you verify, you sign. A private model can still be wrong. It is just privately wrong.

The full breakdown, including who it fits, what the hardware looks like, and how the build runs, is in The Private AI Workup in the downloads above.

The REALTOR's Choice for AI Builds and Systems

You are managing escrows, writing contracts, and showing property. You do not have 40 hours to troubleshoot API keys and webhooks. I build these exact systems for real estate agents and then hand you the keys. I am not a software vendor who read a book about real estate. I have been licensed in California since 1998, I carry the same fiduciary duty and the same Fair Housing exposure you do, and I have trained more than a thousand agents on this. That is why agents call me to build it instead of a developer who has never sat at a kitchen table at 7 at night.

Your AI Employee

Context files, memory, and skills built from your real voice, your splits, your compliance rules, and your farm.

Speed-to-Lead Machine

Intake, scoring, and instant follow-up that filters tire-kickers and pushes high-intent buyers to your phone.

24/7 Voice Receptionist

Sign calls and buyer inquiries answered, qualified, and booked to your calendar while you sleep.

AEO Entity Cleanup

Get your name into the short list the answer engines recommend when someone asks who the best agent in your market is.

Private AI Systems

Your own model on your own hardware, for agents, brokerages, and attorneys. I spec the components, install it, train it on your business, and hand you the keys. I run one in my own residence.

Audit → Blueprint → Build → Handoff  ·  NDA available on every engagement  ·  Built by a licensed REALTOR, not a vendor

And back to her

It is six months later. Same person, same hour, and she is not shopping anymore. She is scared.

The deal is wobbling and she is lying awake doing math about her family. No machine answers that one. An algorithm cannot look a terrified client in the eye at 11 o'clock at night and say I know you are scared, I am standing right here beside you, and I will protect you.

The machine got you into the conversation. It cannot finish it. That is the job. It has always been the job.

AI is not replacing real estate agents. Agents using AI will replace agents who refuse to.

This has always been a relationship business and nothing on this page changes that. All of it protects it. Give away the paperwork. Keep the people.

Book Your Free 15-Minute AI Business Audit

A short Zoom. We look at your CRM, email, and marketing setup, find your biggest bottleneck, and hand you a custom automation blueprint. Zero pressure. The blueprint is yours either way.

Prefer a bookmark? The same calendar lives at BookWithHonor.com

Or call or text me directly: 661-400-1720

Quick Answers

What happens on the 15-minute audit?
We hop on Zoom, look at your current setup, and identify the one bottleneck costing you the most hours. You leave with a step-by-step blueprint whether we ever work together or not.
Do I need to be technical?
No. Start with a paid ChatGPT Plus or Claude Pro account and your phone's voice mode. Everything past that can be built for you.
Will AI replace real estate agents?
No. AI will never replace real estate agents. But AI-powered agents will replace those who refuse to adapt. The relationship is yours. The paperwork is the machine's.
Is my client data protected?
Every build follows the principle of least privilege, human checkpoints sit before anything reaches a client, and I will sign an NDA with you before we start.
Can I paste my client list into ChatGPT?
Be careful even in a paid account. A paid workspace with training turned off is the floor for any client information, but training off does not mean local: the data still travels to the provider, gets processed on their servers, and sits in their logs for a period. Send the minimum the job needs and strip names and account numbers the task does not require. And protected documents never go into any cloud AI tool, free or paid: pre-approval letters, executed contracts, lockbox codes, and anything covered by financial, medical, or credit privacy law. If a job truly requires those, it runs on a local model on your own hardware or it does not happen. This is the single mistake in the whole workflow that can cost you your license.
Is incognito or private mode on ChatGPT, Gemini, or Claude safe for client data?
No. Temporary or incognito chat protects you from your own chat history, not from the company. It keeps the conversation out of your saved history and usually out of training data, but the provider still receives everything, still ties it to your account, and typically retains logs for around 30 days for safety review, sometimes longer under legal holds. Browser incognito adds nothing on top of that. Settings change what they keep, not where it goes. So the rule does not bend: pre-approval letters, executed contracts, lockbox codes, and legally protected personal information never go into any cloud AI tool, free or paid, incognito or not. Everyday client work belongs in a paid workspace with training turned off, with the minimum information the job needs. Truly private means local: a model on your own hardware, or it stays out.
Can I have an AI where protected documents CAN go in?
Yes: a private system, where the model runs on hardware you own inside your own building. No account, no provider, no logs on anyone else's servers. Contracts, client files, and a brokerage's whole transaction history can go in because nothing ever leaves the room. I run one in my own residence on compute I purchased myself, and I build them for agents, brokerages, and attorneys: I spec the right components, install the model and the private document search, train it on your business, and hand your staff the keys. The trade is that local models are less powerful than the big cloud models, so the smart setup is both: cloud for public work, private for everything the cloud must never see.
Does Fair Housing apply to marketing copy that AI wrote?
Yes. It applies to AI-generated marketing copy, to ad targeting, and to anything that screens people. Disparate impact means your intent does not save you. The liability lands on you and your broker, not the software vendor, so your Fair Housing parameters belong in the context file explicitly. The model follows what you said and will never know what you meant.
Is SEO dead now that AI answers everything?
No, and acting like it is will hurt your business. Pew Research found people click a traditional result 8 percent of the time when an AI summary is present versus 15 percent when it is not, and Ahrefs measured the number one spot losing between 34.5 and 58 percent of its clicks under an AI Overview. Ranking still matters, it just pays about half of what it used to. Meanwhile AI referrals are still roughly two tenths of one percent of web traffic. Abandoning SEO for AEO is abandoning the highway for the bike path. Build both.
Why hire a REALTOR to build AI instead of a software developer?
Because the exposure is the whole problem. I have been licensed in California since 1998, I carry the same fiduciary duty and the same Fair Housing liability you do, and I have trained more than a thousand agents on these systems. A developer can wire an API. A developer cannot tell you which of two perfectly legal options is the right one for a specific family in a specific week, or where a workflow quietly creates a disparate impact problem that lands on you and your broker.