Build a real estate AI assistant

Build an AI assistant purpose-built for real estate that handles property inquiries around the clock, schedules showings, qualifies buyers and sellers, and answers market questions — so you can focus on closing deals instead of fielding repetitive questions.

Overview


Goal	Deploy an AI assistant that handles property inquiries, schedules showings, qualifies buyers and sellers, and answers common real estate questions
Time to build	90–120 minutes
Prerequisites	Conversation AI enabled, Voice AI (optional), calendar configured, custom fields created
AI features used	Conversation AI + Voice AI + Calendar + Custom Fields
Difficulty	Advanced

Voice AI agents list showing agents with inbound direction and phone numbers — Voice AI agent setup for real estate with agent name, direction, and phone number

Architecture overview

The real estate AI assistant operates across the entire client journey:

Inquiry capture — A prospect reaches out about a listing via SMS, web chat, phone, or social media ad.
Property information — The bot answers questions about the property using the knowledge base (price, bedrooms, square footage, neighborhood, etc.).
Buyer/seller qualification — The bot asks qualifying questions to determine if the prospect is a serious buyer or seller.
Showing scheduling — Qualified prospects are offered available showing times and booked onto the agent’s calendar.
CRM update — The bot tags the contact, updates custom fields with qualification data, and moves them through the pipeline.
Agent notification — The real estate agent receives an alert with the prospect’s details and showing information.
Follow-up — Automated follow-up sequences nurture prospects who are not ready to buy or sell immediately.

Step-by-step build

Create real estate custom fields

Navigate to Settings > Custom Fields and create these fields on the Contact object:

Field name	Type	Purpose
Buyer or Seller	Dropdown	Options: Buyer, Seller, Both, Investor
Property Interest	Text	Address or MLS number of the property they inquired about
Budget Range	Dropdown	Options: Under $200K,$ 200K– $400K,$ 400K– $600K,$ 600K– $1M,$ 1M+
Pre-Approved	Dropdown	Options: Yes, No, In progress, Cash buyer
Timeline	Dropdown	Options: Immediately, 1–3 months, 3–6 months, 6+ months, Just browsing
Current Agent	Dropdown	Options: No, Yes — under contract, Yes — not under contract
Property Type	Dropdown	Options: Single family, Condo, Townhouse, Multi-family, Land, Commercial
Neighborhoods of Interest	Text	Preferred areas or zip codes
Showing Scheduled	Checkbox	Whether a showing has been booked

The “Current Agent” field is critical. If a prospect already has an agent under contract, your bot should handle that gracefully rather than trying to poach the client.

Set up your CRM pipeline

Create a real estate pipeline with these stages:

New inquiry — Just came in, not yet engaged
Qualification in progress — Bot is qualifying
Qualified buyer — Ready for showings
Qualified seller — Ready for listing presentation
Showing scheduled — Has an upcoming showing
Post-showing follow-up — Attended showing, needs follow-up
Offer stage — Actively making or receiving offers
Under contract — Deal in progress
Closed — Deal complete
Nurture — Not ready now, follow up later

Build the property knowledge base

Navigate to AI Agents > Knowledge Base and add:Active listings (update regularly):

Address, price, bedrooms, bathrooms, square footage
Key features (pool, garage, updated kitchen, etc.)
Neighborhood information (schools, restaurants, commute times)
Open house dates and showing availability
Photos description (the bot cannot show images but can describe them)

General real estate information:

Your service area and specializations
The buying process (step by step for first-time buyers)
The selling process (what to expect)
Financing basics (types of loans, pre-approval process)
Market conditions (average days on market, price trends)
Your team bios and specializations

Update your knowledge base every time you add, remove, or change a listing. Outdated listing information is the fastest way to lose trust with a prospect.

Configure the calendar for showings

Create a dedicated calendar for property showings:

Availability: Set windows when you can do showings (e.g., 10 AM – 6 PM, Tuesday – Saturday)
Duration: 30 minutes per showing (with 15-minute buffer for travel)
Booking window: Allow bookings 1–14 days out
Minimum notice: 4 hours (gives you time to prepare and confirm access)

See creating a calendar for setup details.

Create the Conversation AI bot

Navigate to AI Agents > Conversation AI and create a bot:

Name: “Real Estate Assistant”
Type: Auto-pilot
Channels: SMS, Web Chat, Facebook Messenger, Instagram
Calendar: Assign your showings calendar

See Conversation AI setup for detailed steps.

Create the Voice AI agent (optional)

For phone inquiries, create an inbound Voice AI agent:

Name: “Property Hotline”
Direction: Inbound
Calendar: Same showings calendar
Voice: Choose a warm, professional voice

Assign it to a dedicated property hotline number. See Voice AI phone setup.

Write the bot prompt

Paste the full prompt from the section below. The real estate prompt is more complex than other use cases because it handles multiple prospect types (buyers, sellers, investors) and needs property-specific knowledge.

Configure bot actions

Set up Conversation AI actions:

Update custom fields: Map buyer/seller type, budget, timeline, pre-approval status
Add tags: “buyer,” “seller,” “investor,” “showing-booked,” “pre-approved”
Move opportunity: Advance through the real estate pipeline based on conversation outcome
Book appointment: Connect to the showings calendar for scheduling

Build follow-up workflows

Create these workflows:Post-inquiry nurture (for prospects not ready to act):

Day 1: Send a market report for their area of interest
Day 7: Share a new listing that matches their criteria
Day 14: Check in with a soft question about their timeline
Day 30: Monthly market update
Continue monthly until they re-engage or opt out

Post-showing follow-up:

2 hours after showing: “What did you think of [ADDRESS]? I would love to hear your thoughts.”
Day 2: If positive response, ask if they want to make an offer
Day 2: If negative, ask what they are looking for instead and suggest alternatives

Seller follow-up:

Day 1: Send a comparative market analysis (CMA) for their property
Day 3: Follow up asking if they reviewed the CMA
Day 7: Offer a free listing consultation

See workflow automation overview for configuration details.

Full prompt

You are a real estate AI assistant for [AGENT NAME] at [BROKERAGE NAME]. You help prospective buyers and sellers with property inquiries, schedule showings, and qualify leads.

IDENTITY AND TONE:
- You represent [AGENT NAME], a licensed real estate agent specializing in [AREAS/NEIGHBORHOODS].
- Be professional, knowledgeable, and enthusiastic about real estate without being pushy.
- Use the prospect's name once you learn it.
- Keep SMS messages to 2-3 sentences. Web chat can be slightly longer.

HANDLING PROPERTY INQUIRIES:
- When someone asks about a specific property, provide details from the knowledge base: price, bedrooms, bathrooms, square footage, key features, and neighborhood highlights.
- If you do not have information about a specific property, say: "I do not have the details on that property handy, but [AGENT NAME] can pull up all the information for you. Would you like me to schedule a call or showing?"
- Never make up property details. Only share information from the knowledge base.

BUYER QUALIFICATION FLOW:
1. "Are you currently looking to buy a home, or just exploring your options?"
2. "What areas or neighborhoods are you most interested in?"
3. "What type of property are you looking for?" (single family, condo, etc.)
4. "What is your price range or budget?"
5. "Have you been pre-approved for a mortgage, or are you a cash buyer?"
6. "What is your timeline for purchasing?"
7. "Are you currently working with another real estate agent?"

SELLER QUALIFICATION FLOW:
1. "Are you thinking about selling your home?"
2. "What is the address of the property you are considering selling?"
3. "What is your timeline for selling?"
4. "What is motivating your move?" (upsizing, downsizing, relocation, etc.)
5. "Have you had a recent appraisal or do you know your estimated home value?"
6. "Are you currently working with a listing agent?"

SCHEDULING SHOWINGS:
- For qualified buyers, offer to schedule a showing: "I would love to set up a showing for you. Let me check [AGENT NAME]'s availability."
- Check the calendar and offer 2-3 available times.
- Confirm: "You are booked for a showing at [ADDRESS] on [DATE] at [TIME] with [AGENT NAME]. You will receive a confirmation text with details."

HANDLING THE "ALREADY HAVE AN AGENT" RESPONSE:
- If they say they have an agent under contract: "No worries at all! I hope you find exactly what you are looking for. If things change, feel free to reach out anytime."
- If they say they have an agent but are not under contract: "Understood! We are always happy to help. If you would like to see any of our listings or get a second opinion, just let me know."
- Never disparage other agents or try to convince someone to break an existing relationship.

MARKET QUESTIONS:
- Answer general market questions from the knowledge base (average prices, days on market, trends).
- For specific valuation questions: "For an accurate home valuation, I would recommend a personalized comparative market analysis. [AGENT NAME] can prepare one for you at no cost. Would you like me to set that up?"

BOUNDARIES:
- Never provide specific legal or tax advice — suggest they consult an attorney or accountant.
- Never guarantee property values will increase or that a home will sell for a specific price.
- Never share other clients' personal information or transaction details.
- Never discuss commission rates — say "[AGENT NAME] can discuss that with you directly."
- If a question is outside your knowledge, offer to have the agent follow up personally.

Testing your real estate assistant

Property inquiry test

Ask about a specific active listing. Verify the bot provides accurate details from the knowledge base.

Buyer qualification test

Respond as a serious buyer (pre-approved, immediate timeline, specific budget). Verify the bot qualifies you and offers to schedule a showing.

Seller qualification test

Say you want to sell your home. Verify the bot asks the right questions and offers a CMA or listing consultation.

Showing booking test

Accept a showing invitation. Verify the bot checks calendar availability, books the slot, and sends confirmation.

Already-has-agent test

Say “I already have an agent.” Verify the bot handles it gracefully without being pushy.

Unknown property test

Ask about a property not in the knowledge base. Verify the bot admits it does not have the details and offers to connect you with the agent.

CRM verification

After each test, check the contact record for accurate custom fields, tags, and pipeline stage.

Optimization tips

Update listings in the knowledge base weekly. Stale listings are the number one complaint from real estate prospects interacting with AI bots.
Track which listings generate the most inquiries. Use tags to identify high-interest properties and allocate more marketing budget to them.
Personalize follow-up sequences by property type. A condo buyer has different concerns than a single-family buyer. Segment your nurture sequences accordingly.
Add neighborhood guides to the knowledge base. Buyers often ask about schools, restaurants, commute times, and safety. Rich neighborhood data makes your bot significantly more useful.
Monitor buyer-to-showing conversion. If qualified buyers are not scheduling showings, your prompt may need a stronger call to action, or your available times may not match when buyers are free.
Use the bot for open house follow-up. After an open house, trigger the bot to follow up with attendees who signed in, asking for their impressions and offering private showings.

KPIs to measure success

KPI	Target	Where to find it
Inquiry response time	Under 2 minutes	Conversation AI dashboard
Qualification completion rate	60%+ of inquiries fully qualified	Custom field fill rate
Showing booking rate	30%+ of qualified buyers book a showing	Calendar + tags
Showing attendance rate	80%+ of booked showings attended	Calendar + no-show tracking
Lead-to-close conversion	Track by source and agent	Pipeline reporting
Knowledge base accuracy	Under 5% of questions unanswerable	Transcript review
Seller listing rate	20%+ of qualified sellers list with you	Pipeline reporting

Next steps

Lead qualification

Apply advanced lead scoring techniques to your real estate pipeline.

Appointment scheduler

Enhance your showing scheduling with reminders and no-show follow-up.

After-hours support

Handle property inquiries that come in evenings and weekends.

Prompt engineering

Advanced prompt techniques for real estate-specific conversations.

AI Overview

Agent Studio

Conversation AI

Voice AI

Content AI

Workflow AI

Use Cases

Advanced

AI Trust & Safety

Build a real estate AI assistant

Architecture overview

Step-by-step build

Full prompt

Testing your real estate assistant

Optimization tips

KPIs to measure success

Next steps

Lead qualification

Appointment scheduler

After-hours support

Prompt engineering

​Architecture overview

​Step-by-step build

​Full prompt

​Testing your real estate assistant

​Optimization tips

​KPIs to measure success

​Next steps

Lead qualification

Appointment scheduler

After-hours support

Prompt engineering

Architecture overview

Step-by-step build

Full prompt

Testing your real estate assistant

Optimization tips

KPIs to measure success

Next steps