Messy capture
Talked to the seller. Need to follow up. They seemed nervous. Probably need a new plan.
Free class · Systems and automation
How to build everyday AI workflows for real estate.
Most AI output breaks before the model ever sees it, in a vague note or a handoff nobody checked. This class teaches the communication skill underneath every workflow, then shows you how to run a real one from intake to reviewed deliverable.
Ryan Wanner, founder of AI AccelerationInnovation Lab North, Compass Lakeside, HendersonvilleEvery class on this page is free to read and watch. No account needed.
What you'll learn.
- 01Why communication, not prompting, is the skill underneath every AI workflow: human to human, human to AI, AI to AI, and chains of all three.
- 02How to turn a messy conversation into AI-ready input with the Communication Diamond.
- 03A three-question signal check for anything a person or a model tells you.
- 04What a good spec for AI includes, and why you don't need context cards to start.
- 05How to make one AI check its own work before it hands off to the next step.
- 06How to run a pipeline from trigger to reviewed deliverable, and how to find where a broken one failed.
Part 1
Communication is the skill under every workflow.
Call it prompting and you shrink the skill too far. Prompting is one lane. The real question is whether your intent can move from a person, into an AI, into another AI or tool, and back to a person without losing the point.
- 01
Human to human
Client calls, showing recaps, team asks, lender check-ins. You're probably already good at this one.
- 02
Human to AI
Your specs, context, guardrails, audience, intent, and where the output should go.
- 03
AI to AI
One model or agent makes something, checks it, and passes it to the next step, often before you see it.
- 04
Chains
Human to AI to AI to human to tool. Useful work is becoming a chain, not a single chat. Meaning has to survive every link.
Capture the conversation before it disappears.
AI-ready capture
- Outcome: seller chooses a pricing plan
- How: compare three pricing scenarios
- When: before Friday listing prep
- Motivation: reduce days-on-market risk
- Constraint: missing updated comps
- Owner: me, until I hand it to the seller
The second version is the Communication Diamond. Give AI those six pieces and it can file the note, create the task, draft the follow-up, or tell you exactly what's missing instead of pretending.
- 01
What: the outcome
What needs to happen, in one sentence the receiver could repeat back. The only good process is one that supports an outcome.
- 02
How
The steps, example, or instruction the doer needs to get there.
- 03
When
The due date, and what depends on that timing.
- 04
Motivation
Why it matters to you, to them, or to the client. What keeps the doer moving.
- 05
Constraint
What's blocking it: a missing fact, decision, resource, access, or approval.
- 06
Who owns it
Who does the work, and who owns getting it done.
The format isn't sacred. Repeatability is the point. Adjust the labels to how you talk and the CRM you use, but give the AI roughly the same categories every time. When the shape stays the same, the AI has fewer gaps to fill, and fewer gaps means less invention.
Turn notes into a Communication Diamond
Here are my notes from a conversation: [PASTE YOUR MESSY NOTES, REDACTED] Turn these into a Communication Diamond: - Outcome: what needs to happen, in one sentence - How: the steps or instruction needed - When: the due date and what depends on it - Motivation: why it matters and to whom - Constraint: what is blocking it or missing - Owner: who does the work and who owns completion Mark anything my notes don't answer as [MISSING] and ask me about it. Do not fill gaps with guesses. Then tell me the right next artifact: CRM note, task, calendar item, client follow-up, or missing-information checklist.
Part 2
Filter it, spec it, then check every handoff.
The signal check.
Before a claim becomes a prompt, a task, or a client message, run three questions. It works on a colleague's tip and on an AI answer alike.
What does this actually mean? If you can't restate it plainly, it isn't ready to act on.
How do you know? Ask AI for sources and links, then open them. "Someone told me" isn't evidence.
So what? Does this move the business forward, and what would you do differently because of it?
What a good spec for AI includes.
A strong ask is more than a request. When output sounds robotic, don't rewrite the whole prompt. Find the missing slot.
- 01
The finished output
What it should look like: format, quality bar, and where it's going. A client text, a CRM note, a listing email.
- 02
What not to do
Banned phrases, compliance limits, style misses, the no-go list.
- 03
What you already tried
What failed before, so the model doesn't repeat it.
- 04
Business context
Your market, your clients, your tools, how you actually work.
- 05
Audience and intent
Who reads it, whose voice is speaking, and what they should do or decide next.
- 06
The missing-info question
Ask the model what else would help before it writes. Let it ask instead of guess.
Unbundle vague words
"Make it professional." "Make it clearer." "Make it better." Bundled words don't tell anyone what to do. Break them into behavior: which part is wrong, what should happen instead, and how you'll know it's fixed.
No cards? Quick context
Context cards are useful building blocks, not a prerequisite. Without them, paste who you are, the situation, the facts you trust, the audience, the constraints, and the output shape. Load only what this workflow needs. More context isn't automatically better.
Make AI check its work before it hands off.
AI-to-AI is where drift compounds, because several steps can run before you see anything. Say you never want the phrase "dream home." Put it in the guardrails, and the validation step searches the draft, flags every instance, and redoes the job before passing it on.
AI-to-AI validation loop
Before passing this output to the next step, validate it against the operator criteria. Outcome: - What was this step supposed to produce? Handoff document: - What format does the next AI, tool, or human expect? - What required fields must be present? - What evidence or source material must travel with it? - What should be forbidden or marked [review]? Validation: - Does it match the requested output? - Does it include the required sections or fields? - Did it violate a guardrail? - Is anything missing that the next receiver expects? If any check fails: - Name the failure. - Resubmit the job with the missing requirement. - Do not pass the artifact forward until it passes.
Part 3
Run pipelines, not prompts.
"Write me a listing description" depends on what you remembered to paste and whether you catch the mistakes. A pipeline starts where the business starts and ends with approved work. Here's the listing launch, the one the class built live.
- 01
Trigger
The real event that starts the work: a seller intake email, listing appointment notes, an inspection report, a new buyer inquiry.
- 02
Context filter
Pull out the verified facts, the missing facts, the strongest angles, the constraints, and the claims you can't make. Before anything gets written.
- 03
Production
Now the AI drafts: listing remarks, social assets, the email, the seller update. Every draft works only from the filtered facts.
- 04
Validation
Check every factual claim, Fair Housing risk, MLS and brokerage rules, voice drift, disclosure needs, and anything that sounds like hype.
- 05
Human approval
You edit, fill gaps, reject weak sections, and only then move the work to the MLS, email, social, print, or CRM.
- 06
Final deliverable
One folder with the source notes, the context, the approved copy, and the client communication record.
If the intake email says "renovated kitchen" but not when, the output can't claim a new kitchen. If the photos show an empty room, a virtual staging plan needs disclosure and a realism check. The pipeline protects you from confident garbage. You're the licensed human. The AI is the assistant.
Other pipelines worth building.
- Buyer consultation: lead inquiry to consult plan, search criteria, follow-up, and CRM note.
- Pricing narrative: subject property and comps to a seller-facing pricing memo.
- Seller update: showing feedback and activity to a pattern summary and next recommendation.
- Inspection response: the report to options, a client explanation, and a repair request.
- Sphere reactivation: a CRM export to segments, an outreach plan, and a next-action queue.
- Daily deal desk: inbox, calendar, and CRM notes sorted into urgent, this week, FYI, and missing context.
When a pipeline breaks, find the station.
Bad output doesn't automatically mean missing context. Start with the source, then work down.
Source
The facts were weak, stale, private, or too vague.
Context
Notes conflicted, overrode each other, or pointed at the wrong audience.
Spec
The format, destination, quality bar, or definition of done was unclear.
Sequence
The workflow skipped extraction, review, or approval.
Tool
The wrong model or app for the job.
Review
No pass-or-fail check before the output left the chat.
Execution
The AI had enough and still did bad work. Reject it, split the task, or switch tools.
Don't try to automate your whole business. Pick one pipeline, write down its trigger, sources, review gate, and deliverable today, run it on live work tomorrow, and save the whole chain: input, filter, output, edits, approval, and final artifact.
Free with a student account
Get the workbook and prompts, free.
Create a free student portal account to get the workbook, prompts, templates, and downloads for every class.
Create my free account →Already a student? The same link signs you in.
- The Communication + Everyday AI Workflows workbook
- Communication Diamond, signal check, and unbundling templates
- Human-AI spec and AI-to-AI handoff contract templates
- Quick Context Sheet, no context cards required
- Four pipeline recipes: daily deal desk, listing launch, pricing narrative, buyer consultation
- The Everyday AI Workflows skill card for your AI
Questions agents ask.
What is an AI workflow for a real estate agent?
A workflow is a chain, not a prompt. It starts with a real trigger like a seller intake email, filters the facts, has AI draft the output, runs a review, and ends with an approved deliverable such as a listing launch kit or a CRM note. The AI step is one station in the chain.
Why does my AI output break even when my prompt looks good?
Often the prompt isn't the problem. Run a failure review: was the source material weak, did the context conflict, was the spec unclear, did the workflow skip a step, was it the wrong tool, was there no review, or did the AI just fail? Only one of those is missing context.
Do I need context cards before I can use these workflows?
No. Context cards speed up repeated work, but they aren't the ticket in. Without them, write quick context: who you are, the situation, the facts you trust, the audience, the constraints, and the output you want. Turn it into a reusable card later if the workflow keeps repeating.
How do I stop one AI step from passing bad work to the next?
Add a validation step before every handoff. Tell the model what the step was supposed to produce, what the next receiver expects, and what's forbidden, such as a phrase you never use. If any check fails, it names the failure and redoes the job before passing anything forward.
Which workflow should I build first?
The one sitting on your desk this week. Good first choices are a listing launch, a daily deal desk that sorts your inbox and calendar, a buyer consultation plan, or a pricing narrative. Run it on live work, not a made-up example, and save the whole chain.
Keep going.
Systems and automation
AI Automations for Real Estate Agents
Write any recurring job as trigger, middle, exit condition, and iterations, so AI drafts on a schedule and you keep the send button.
Systems and automation
Visual AI for Real Estate: Photos, Screenshots, and Documents
Show AI the screen, document, photo set, or voice note, make it split what it sees from what it can't confirm, then verify.
Systems and automation
How to Build an AI Operating System for Real Estate
Set up one inbox, a short context file about you, clear homes with their own instructions, a system map, and a filing log any AI can work from.
Library
Every free class
The whole library of recorded classes, by topic.
Research sources
Check the current rules, capabilities, and source basis.
- 01OpenAI: A practical guide to building agents ↗
First-party guidance on when agents fit, tools, instructions, guardrails, and human intervention.
- 02NIST: AI Risk Management Framework Core ↗
The Govern, Map, Measure, and Manage framework, including training, intended-use, human-oversight, testing, and documentation outcomes.
- 03HUD: Fair Housing guidance for digital advertising ↗
Government guidance on Fair Housing risk in digital housing advertising and delivery.
- 04NAR: Artificial Intelligence in Real Estate ↗
NAR's current overview of AI use in real estate and policy priorities around privacy, Fair Housing, and copyrighted listing content.
Product access, limits, local listing rules, platform policies, and laws can change. Re-check the linked authority before important or public work.
Keep building
Put AI to work in your real estate business.
AI Acceleration: Real Estate is a private, self-directed curriculum for practical AI work across content, communication, lead follow-up, CRM, client research, operations, and relationship management. Build it. Review it. Keep what works.
Explore the Skool community ↗Follow Ryan + AI Acceleration
No ranking, revenue, closing, or time-saving promise. The classes teach the work and the review systems around it.
Record: innovation-lab-public-classes. Approved record hash: add45e5ab0ee95b80ad70a4ecb50c8b8317c6997119c13fd8a727225815aff73. Claims: public-classes-free-to-watch, public-classes-portal-unlock, public-classes-taught-live.